sign in sign up
<<
Home , Opioid receptor

The Effect of μ- Polymorphisms on

Receptor Signalling Systems

Alisa Knapman (BMedSci Hons)

Faculty of Science

Australian School of Advanced Medicine

Macquarie University

2014

A thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy

Table of Contents

Chapter 1: Introduction 1

1.1 Opioid Receptors 2

1.1.1 MOP 4

1.1.2 DOPr 5

1.1.3 KOPr 5

1.1.4 NOPr 6

1.2 Opioid receptors are coupled receptors 7

1.3 MOPr Structure and Signalling 8

1.4 MOPr Effectors 9

1.4.1 MOPr inhibition of 9

1.4.2 MOPr activation of potassium channels 11

1.4.3 MOPr inhibition of calcium channels 12

1.4.4 MOPr release of intracellular calcium 13

1.4.5 MOPr and the MAP kinase cascade 14

1.4.6 Other MOPr effectors 15

1.4.7 MOPr desensitisation and endocytosis 16

1.5 MOPr Ligands 18

1.5.1 Endogenous MOPr ligands 18

1.5.2 Exogenous MOPr ligands 18

1.5.3 affinity and efficacy 19

1.6 Ligand-biased signalling at MOPr 20

1.7 OPRM1 Polymorphisms 22

1.8 MOPr variants and differential signalling 25

i Chapter 2: Review Article 27 “Cellular signalling of non-synonymous single nucleotide polymorphisms of the human μ-opioid receptor”

Hypotheses and Aims 77

Methods - Overview 78

Chapter 3: Research Article 79 “A real-time, fluorescence-based assay for measuring μ-opioid receptor modulation of adenylyl cyclase activity in Chinese hamster ovary cells”

Chapter 4: Research Article 107 “A continuous, fluorescence-based assay of μ-opioid receptor activation in AtT-20 cells”

Results - Overview 135

Chapter 5: Research Article 137 “ signalling is compromised at the N40D polymorphism of the human μ-opioid receptor in vitro”

Chapter 6: Research Article 185 “ The A6V polymorphism of the human μ-opioid receptor negatively impacts signalling of and endogenous in vitro”

Chapter 7: Research Article (in preparation) 233 “Mu-opioid receptor polymorphisms differentially affect receptor signalling via adenylyl cyclase inhibition and ERK1/2 phosphorylation”

ii Chapter 8: General Discussion and Summary 263

References 275

Appendix: Book Chapter 329 “Fluorescence-Based, High-Throughput Assays for µ-Opioid Receptor Activation using Membrane Potential Sensitive Dye”

iii List of Figures 2.1 Naturally occurring non-synonymous OPRM1 variants 35 3.1 Stimulating adenylyl cyclase hyperpolarises CHO cells 87 3.2 Opioids inhibit forskolin stimulated membrane hyperpolarisation 89 in CHO cells 3.3 activators mimic the effects of forskolin 93 3.4 Forskolin induced hyperpolarisation is dependent on extracellular 95 K concentration 4.1 Example traces of fluorescent signal in the membrane potential assay 114 4.2 Trace of fluorescent signal illustrating reversal of DAMGO stimulated 117 decrease in signal by MOPr antagonist 4.3 DAMGO and morphine stimulated hyperpolarisation is PTX sensitive 120

4.4 Decreasing [K]Ex increases maximum change in fluorescent signal 120 4.5 Nigericin causes maximum membrane hyperpolarisation 121 4.6 DAMGO, morphine, buprenorphine and concentration- 121 response curves for GIRK activation 4.7 Desensitisation of MOPr signalling in AtT-20 cells 123 5.1 Saturation binding curve of [3H]DAMGO in intact CHO-MOPr cells 149 5.2 DAMGO inhibits adenylyl cyclase and activates ERK1/2 in CHO cells 153 expressing MOPr-WT and MOPr-N40D 5.3 Endogenous opioids inhibit adenylyl cyclase and activate ERK1/2 in 155 CHO cells expressing MOPr-WT or MOPr-N40D 5.4 Buprenorphine inhibits adenylyl cyclase and activates ERK1/2 less 157 effectively in CHO cells expressing MOPr-N40D 5.5 , and inhibit adenylyl cyclase and activate 159 ERK1/2 in CHO cells expressing MOPr-WT or MOPr-N40D 5.6 DAMGO causes membrane hyperpolarisation in AtT-20 cells expressing 164 MOPr-WT 5.7 Buprenorphine is less potent for GIRK activation in AtT-20 cells 167 expressing MOPr-N40D 6.1 Saturation binding curve of [3H]DAMGO in intact CHO-MOPr-WT 196 and CHO-MOPr-A6V cell 6.2 DAMGO inhibits adenylyl cyclase and activates ERK1/2 in CHO cells 199 expressing MOPr-WT and MOPr-A6V

iv

6.3 Morphine, buprenorphine and pentazocine inhibition of adenylyl cyclase 203 and activation of ERK1/2 is compromised in CHO cells expressing MOPr-A6V 6.4 Endogenous opioid inhibition of adenylyl cyclase and activation of 207 ERK1/2 is significantly affected by the A6V variant in CHO cells expressing MOPr 6.5 Fentanyl, methadone and oxycodone inhibition of adenylyl cyclase 211 and activation of ERK1/2 is significantly affected by the A6V variant in CHO cells expressing MOPr 6.6 DAMGO causes membrane hyperpolarisation in AtT-20 cells expressing 213 MOPr-WT 6.7 Morphine activates GIRK less effectively in AtT20 cells expressing 215 MOPr-A6V 7.1 DAMGO, morphine, buprenorphine and pentazocine inhibit adenylyl 245 cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr variants 7.2 Endogenous opioids inhibit adenylyl cyclase and activate ERK1/2 247 in CHO cells expressing MOPr-WT or MOPr variants 7.3 Methadone, fentanyl and oxycodone inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr variants 251 7.4 High concentrations of methadone and fentanyl cause membrane 252 depolarisation in CHO-MOPr-L85I cells

v List of Tables

2.1 Summary of non-synonymous MOPr variants 38 2.2 Summary of key findings about MOPr SNP signalling 48 - 51 4.1 Potencies and efficacies of the range of structurally distinct 119 opioid ligands tested using the membrane potential assay 5.1 Summary of opioid efficacy and potency in assays of AC 160 inhibition in CHO cells expressing MOPr-WT and MOPr-N40D 5.2 Summary of opioid efficacy and potency in assays of ERK1/2 161 phosphorylation in CHO cells expressing MOPr-WT and MOPr-N40D 5.3 Summary of opioid efficacy and potency in assays of GIRK 168 activation in in CHO cells expressing MOPr-WT and MOPr-N40D 6.1 Summary of opioid efficacy and potency in assays of AC 200 inhibition in CHO cells expressing MOPr-WT and MOPr-A6V 6.2 Summary of opioid efficacy and potency in assays of ERK1/2 208 phosphorylation in CHO cells expressing MOPr-WT and MOPr-A6V 6.3 Summary of opioid efficacy and potency in assays of GIRK 216 activation in in CHO cells expressing MOPr-WT and MOPr-A6V 7.1 Surface receptor expression and DAMGO Kd for MOPr 242 variants in CHO cells 7.2 Forskolin and PMA responses in MOPr variants 242 7.3 Summary of opioid efficacy and potency in assays of AC 254 inhibition in CHO cells expressing MOPr-WT and MOPr variants 7.4 Summary of opioid efficacy and potency in assays of ERK1/2 255 phosphorylation in CHO cells expressing MOPr-WT and MOPr variants

vi Abbreviations

4-AP 4-aminopyridine aa amino acid

AC adenylyl cyclase

ACTH adrenocorticotropic hormone

Akt protein kinase B

β-CNA β-

β-FNA β-funal trexamine

BSA bovine serum albumin

CaM calmodulin cAMP cyclic adenosine monophosphate

CaMKII Ca2+/calmodulin-dependent protein kinase II

CHO Chinese Hamster Ovary

CRC concentration-response curve

CRE cAMP response element

DAMGO [D-Ala2, N-MePhe4, Gly-ol]-

DMEM Dulbecco’s Modified Eagle Medium

DOPr δ-opioid receptor

DPDPE [2-Dpenicillamin, 5-Dpenicillamin]-enkephalin

ECL extracellular loop

EGFR epidermal growth factor receptor

ELISA -linked immunosorbent assay

EM1 1

EM2 endomorphin 2

EPAC exchange signal activated by cAMP

vii ERK extracellular signal regulated kinase

FSK forskolin

GIRK G protein gated, inwardly rectifying potassium

channel

GLR3 glycine receptor α3

GPCR G protein couple receptors

GRK G protein coupled receptor kinase

HBSS Hanks balanced salt solution hMOPr human µ-opioid receptor

HPA hypothalamic-pituitary-adrenal

ICa voltage gated Ca channels

ICL intracellular loop

IP3 inositol (1,4,5) triphosphate

JNK Jun N-terminal kinase

KO knock-out

KOPr κ-opioid receptor

M-6-G morphine-6-glucuronide

MAPK mitogen activated protein kinase

MOPr µ-opioid receptor mRNA messenger ribonucleic acid

N/OFQ /orphanin FQ

NMDA N-methyl-d-aspartate

NOPr nociceptin/orphanin FQ receptor

NR1 NMDA receptor 1

OPRD1 opioid receptor δ 1

viii OPRK1 opioid receptor κ 1

OPRL1 opioid receptor-like 1

OPRM1 opioid receptor mu 1 p38MAPK p38 mitogen activated kinase

PAG periaqueductal grey pDyn pENK pERK phosphorylated ERK1/2

PI3K phosphatidylinositol-3 kinase

PKA protein kinase A

PKC protein kinase C

POMC

PLC phospholipase C

PTX pertussis toxin

SNP single nucleotide polymorphism

STAT5 signal transducer and activator of transcription 5

TEA tetraethylammonium

TM transmembrane

TRPV1 transient receptor potential cation channel, subfamily

V, member 1

ix

x Summary

There is significant variation in individual response to opioid drugs, one cause of which is likely to be polymorphisms on the opioid receptors themselves. The μ-opioid receptor

(MOPr) is the primary site of action for most opioids. Previous studies have identified a number of naturally occurring single nucleotide polymorphisms (SNPs) in the coding region of MOPr. The A118G SNP (N40D), present at allelic frequencies ranging from 10 – 50%, has been associated with diverse phenotypic effects as well as differences in receptor signalling in vitro, with little consistency between studies. Few studies have examined the consequences of other MOPr polymorphisms on receptor function, or potential ligand and pathway specific effects. In this study, the relative potency and efficacy of a range of clinically important and/or structurally distinct opioid ligands were assessed in Chinese hamster ovary (CHO) cells and mouse pituitary neuroblastoma (AtT-

20) cells stably transfected with human wild type MOPr and 5 naturally occurring MOPr variants, N40D, A6V, L85I, R260H and R265H. MOPr surface expression levels were similar for all variants examined. MOPr activation was measured using a membrane- potential assay of adenylyl cyclase (AC) inhibition, a whole-cell ELISA of extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation, and a membrane-potential assay of

G protein-activated potassium channels (GIRKs). In cells expressing MOPr-N40D, buprenorphine inhibition of AC and stimulation of ERK1/2 was significantly reduced, with

GIRK activation unaffected. In cells expressing MOPR-C17T, buprenorphine signalling was abolished, with a loss of potency of morphine and other ligands. AC inhibition via non-morphinan opioids was enhanced at L85I, while a significant loss of potency for many opioids was observed at R260H. There were minor alterations in the signalling profile of

R265H. These results suggest that MOPr SNPs have the potential to significantly alter receptor function, and may contribute to the individual variability in response to opioid observed clinically. xi Declaration of originality

I hereby declare that the work presented in this thesis has not been submitted for a higher degree to any other university or institution. To the best of my knowledge this submission contains no material previously published or written by another person, except where due reference is stated otherwise. Any contribution made to the research by others is explicitly acknowledged at the beginning of each Chapter.

Approval from the Institutional Biosafety Committee was obtained to create the cell lines expressing the human MOR variants. These are simple genetically modified organisms that are classified as exempt dealings by the Gene Technology regulator. Biosafety approval numbers are IBC REF: 5201000742, 5201200023.

Alisa Knapman

Australian School of Advanced Medicine

Faculty of Human Science

Macquarie University

30th August, 2014

xii Acknowledgements

First and foremost, many thanks to my supervisor Professor Mark Connor for providing me with the opportunity to work in his laboratory, for his unwavering support and encouragement, for all his time and patience in dealing with my unending emails and niggling questions, his guidance at every step along the way, and for never letting me give up! Thank you for being a wonderful supervisor, I couldn’t have done this without you.

I am grateful to my co-supervisor Professor Macdonald Christie for providing AtT-20 cells used in development of the GIRK assay, for allowing me to perform experiments in his laboratory, and to Yan Ping Du for the technical assistance and training in radioligand binding assays.

Much appreciation to Marina Santiago of the Connor lab who performed GIRK channel activation assays and transfected AtT-20 cells with MOPr variants. Her assistance and generosity of time and effort were invaluable.

Thank you to Cheryl Handford of the Department of Pharmacology at Sydney University for her patient technical assistance in performing plasmid preparation of μ-opioid receptor variants. Thank you also to Courtney Breen from the Glass laboratory in Auckland for providing a protocol for fluorescent detection of phosphorylated ERK1/2.

Many thanks to Prof Peter McIntyre for the generous gift of TREX FlpIn-CHO cells.

This work was supported in part by NHMRC Project Grant 1011979 to Macdonald Christie and Mark Connor "The mechanisms responsible for tolerance at the μ opioid receptor". I was the recipient of MQRes scholarship and a top-up allowance from the Australian School of Advanced Medicine.

Finally, to all the past and present members of the Connor lab, Marika Heblinski, Rozhin Ashgari, Jordyn Stuart, Dharmica Mistry, Amelia Edington, Will Redmond and Max Camo, as well as many others at ASAM – thank you for all the support, fun and laughs, loved working with you all!

xiii

xiv Chapter 1

INTRODUCTION

Opioid analgesics are the most widely prescribed drugs in the treatment of moderate to severe and include drugs such as morphine and oxycodone. Despite their powerful analgesic effects, opioids are associated with a number of side effects including respiratory , constipation, nausea, itching and (Matthes et al., 1996). Furthermore, prolonged use of opioid analgesics can cause the development of tolerance, as well as dependence in some individuals, posing a major challenge in the management of long-term or chronic pain (Moore & McQuay, 2005; Dahan et al., 2010; Noble et al., 2010). Over time, increasing doses of opioids become necessary to maintain analgesia, or maintain in the case of opioid abusers (Buntin-Mushock et al., 2005). The increased toxic side effects associated with escalating doses of opioids can reach unacceptable levels, leading to inadequate pain relief or overdose (Corbett et al., 2006). Despite these limitations, morphine and other opioids are the most effective medications currently available to treat the majority of severe and chronic pain conditions, and remain the gold standard in pain relief (Hanks et al., 2001). As such, the development of improved opioid analgesics with increased analgesic effects, reduced adverse effects and less ability to promote tolerance is an area of intense research.

The degree to which tolerance and adverse effects are observed in individuals varies significantly, limiting the ability of physicians to determine the appropriate dosage to balance adequate pain relief with minimal side effects and tolerance. It is likely that genetic factors contribute to differences in these parameters between individuals (Lotsch &

Geisslinger, 2005; Skorpen et al., 2008). The μ-opioid receptor (MOPr) is the primary site

1 of action for most clinically important opioid drugs, including morphine, and thus differences in MOPr function arising from genetic polymorphisms between individuals may contribute to variation in opioid response (Lotsch & Geisslinger, 2005; Mague &

Blendy, 2010). An understanding of how MOPr polymorphisms affect receptor function may assist in predicting an individual’s response to opioid drugs, and potentially lead to more rational and individualized pharmacotherapies, maximizing benefits and minimizing harm.

1.1 Opioid Receptors

Opium and its derivative have been used for centuries to relieve pain and alter mood. Morphine, the prototypic opioid , was isolated from the poppy in

1806, and has since been effectively used for pain relief (van Ree et al., 1999). Specific binding sites were proposed in 1954 (Beckett et al., 1954), and first demonstrated in mammalian tissue in 1973 using radioligand binding studies (Pert & Snyder, 1973; Simon et al., 1973; Terenius et al., 1973). Three opiate receptor subtypes have since been cloned,

µ-opioid receptor (MOPr), κ-opioid receptor (KOPr), and δ-opioid receptor (DOPr), and a fourth opioid-related nociceptin/orphanin FQ receptor (NOPr) (Martin et al., 1976; Lord et al., 1977; Meunier et al., 1995). It has been proposed that these major opioid receptor classes can be further divided into receptor subtypes, however there is little convincing evidence for their existence (Alexander et al., 2013; Connor & Kitchen, 2006; Dietis et al.,

2011).

Opioid receptors, together with their endogenous opioid ligands, make up a complex neuromodulatory system that plays a major role in the control of pain, motivation and reward. Activation of opioid receptors mediates a range of physiological responses such as analgesia, euphoria, feeding, immunomodulation, hormone release and stress responses, as

2 well as some autonomic functions including respiration, gastrointestinal motility, blood pressure and thermoregulation (Boom et al., 2012; Brock et al., 2012; Drolet et al. 2001).

MOPr, DOPr, KOPr and NOPr are differentially expressed throughout the central nervous system (CNS, Mansour et al., 1995, Peckys & Landwehrmeyer, 1999; Reinscheid et al.,

2000), the (Sternini, 2001), and peripheral sensory and autonomic nerves (Coggeshall et al., 1997). Some regions express all receptor types, such as the striatum and dorsal horn of the , although not necessarily in the same neurons, while other regions express an abundance of one receptor type, for example KOPr in the (Mansour et al., 1995). There appear to be significant species differences in opioid receptor distribution (Blackburn et al. 1988; Peckys & Landwehrmeyer, 1999;

Rothman et al., 1992; Sharif & Hughes, 1989). However, rat and human opioid receptor distribution appears to be homologous for many regions (Peckys & Landwehrmeyer,

1999).

The four genes that encode the opioid receptor subtypes are OPRM1 for MOPr, OPRD1 for DOPr, OPRK1 for KOPr and OPRL1 for NOPr (Pawson et al., 2014). All four genes are share a high degree of sequence similarity, and are approximately 60% identical to each other (Law et al., 2000; Satoh & Minami, 1995). The most highly conserved regions are the intracellular loops (85 – 100% identity) and transmembrane domains (75% identity), with greater diversity found within the extracellular N-terminus, the extracellular loops and the intracellular C-terminus (Law et al., 2000; Reinscheid et al., 2000; Satoh &

Minami, 1995; Wei & Loh, 2011). Opioid receptors are highly conserved between species, with human opioid receptors 85-90% identical to their rodent counterparts in terms of protein sequence (Kieffer, 1995). Studies in knockout (KO) mice and rats have helped to elucidate the physiological role each of the classes of opioid receptors, and determine ligand selectivity in vivo. Although only one gene for each receptor type has been

3 identified, the diversity of effects arising from opioid receptor activation led to the hypothesis of multiple receptor subtypes. However studies in mice show abolition or a significant reduction of all opioid receptor effects when the corresponding gene is inactivated (Kieffer, 1999; Dietis et al., 2011).

1.1.1 MOPr

MOPr is the main site of action for morphine and other clinically important opioid analgesics, as demonstrated by a number of studies in OPRM1 knockout mice. In mice that do not express MOPr, morphine has no analgesic effect, regardless of the dose or route of administration (Fuchs et al., 1999; Matthes et al., 1996; Matthes et al., 1998; Sora et al.,

1997; Sora et al., 1999). Furthermore, morphine administration does not cause any of morphine’s associated adverse effects, nor generate a reward response or the development of tolerance in knockout mice (Becker et al., 2000; Contarino et al., 2002; Matthes et al.,

1996). MOPr knockout mice also show reduced reward in response to , and in some cases an aversion (Kieffer & Gaveriaux-Ruff, 2002). In the absence of drug, knockout mice demonstrate an increased sensitivity to certain types of pain, and some behavioural differences such as increased and social withdrawal have been observed (Kieffer &

Gaveriaux-Ruff, 2002; Martin et al., 2003). MOPr is the most highly expressed of all the opioid receptor types, and is widely distributed throughout the CNS, with intense labelling in the , striatum, and periaqueductal grey (PAG) regions of rat and human brain (Mansour et al., 1994; Mansour et al., 1995; Peckys & Landwehrmeyer,

1999), and the superficial laminae of the spinal cord (Mansour et al., 1994; Arvidsson et al., 1995). MOPr can be expressed on excitatory and inhibitory neurons, both pre- or post- synaptically, and the net effect of MOPr activation may be either an inhibitory effect on ascending pain pathways, or a disinhibitory effect on descending analgesic pathways

(Taylor, 2009).

4 1.1.2 DOPr

Although the existence of MOPr and KOPr had already been proposed based on experiments in vivo, DOPr was the first opioid receptor to be isolated, and was extracted from mouse vas deferens (Lord et al., 1977). DOPr has been associated with several physiological functions including analgesia, locomotion, reproduction and mood (Chung &

Kieffer, 2013; Gaveriaux-Ruff & Kieffer, 2011; Wylot et al., 2013). DOPr knockout mice show no or subtle changes in sensitivity to acute thermal, chemical or mechanical pain, but show significantly increased neuropathic and inflammatory pain, suggesting endogenous

DOPr activation may modulate chronic pain (Gaveriaux-Ruff et al., 2008; Gaveriaux-Ruff et al., 2011; Nadal et al., 2006; Pradhan et al., 2013). Activation of DOPr with selective such as DPDPE ([2-Dpenicillamin, 5-Dpenicillamin]-enkephalin) and

II result in analgesia, however the activity of DPDPE and deltorphin II are maintained, reduced or abolished in DOPr knockout mice depending on the assay and route of administration (Scherrer et al., 2004; Zhu et al., 1999). DOPr selective agonists also show reduced activity in MOPr knockout mice (Fuchs et al., 1999; Matthes et al., 1998), suggesting a functional interaction of DOPr and MOPr in vivo (Gomes et al., 2011;

Stockton & Devi, 2012). DOPr knockout mice have been shown to exhibit increased anxiety and depressive-like behaviours (Filliol et al., 2000; Roberts et al., 2001; Lutz &

Kieffer, 2013), changes in endocrine function (Wylot et al., 2013), and altered learning and reward behaviours (Le Merrer et al., 2013; Nieto et al., 2005).

1.1.3 KOPr

Activation of KOPr is associated with many effects including analgesia, immunomodulation, stress response and (Liu-Chen, 2004). Studies in mice lacking the OPRK1 gene showed no change in sensitivity to mechanical or thermal pain, but a significantly increased response to visceral pain (Simonin et al., 1998). The

5 analgesic, hyperlocomotor and aversive effects of the KOPr selective U50-488H were absent in OPRK1 knockout mice (Simonin et al., 1998). KOPr knockout mice also display an increase in generation of antibodies in response to immunisation (Gaveriaux-

Ruff et al., 2003), and a decrease in self-administered alcohol (Kovacs et al., 2005).

Activation of KOPr causes dysphoria, particularly in response to stress, and decreases reward response stimulated by other non-opioid drugs such as cocaine and ethanol

(Charbogne et al., 2014, Land et al., 2008, McLaughlin et al., 2003). KOPr agonists are therefore of particular interest as potential therapeutics for the treatment of addiction (Wee

& Koob, 2010), and their ability to produce analgesia with a low potential for abuse (Wang et al., 2010).

1.1.4 NOPr

NOPr has little affinity for ligands of the classical opioid receptors, and as such was characterized as belonging to the opioid receptor family a number of years after MOPr,

DOPr and KOPr (Meunier et al., 1995). Nociceptin, a selective endogenous NOPr agonist, stimulates a wide and often conflicting range of responses, including hyperalgesia, analgesia, allodynia, as well as anxiolytic and locomotor effects, modulation of other opioid responses and changes in learning and memory (Meis, 2003). Mice lacking the

OPRL1 gene display both increased, decreased and unchanged sensitivity to pain, depending on the assay conditions (Meunier et al., 1997). Studies in NOPr knockout mice have also demonstrated both a suppressive and facilitative role for the receptor in reward and addiction to opioids and other drugs of abuse (Marquez et al., 2008a; Marquez et al.,

2008b; Rutten et al., 2011; Sakoori & Murphy, 2007), changes in locomotion and mood behaviours (Rizzi et al., 2011) and both increased and unchanged anxiolytic effects

(Gavioli et al., 2007). The conflicting nature of many of these findings highlights the

6 importance of specific neuronal circuits in the physiological effects of NOPr activation

(Heinricher, 2003).

1.2 Opioid receptors are G-protein coupled receptors

MOPr and the other opioid receptors were cloned in the 1990s and characterized as Type A or -like G-protein coupled receptors (GPCRs, (Bunzow et al., 1994; Chen et al.,

1993; Evans et al., 1992; Kieffer et al., 1992; Li et al., 1993; Meng et al., 1993; Mollereau et al., 1994; Thompson et al., 1993; Yasuda et al., 1993). More recently, crystal structures have been solved for MOPr (Manglik et al., 2012), DOPr (Granier et al. 2012), KOPr (Wu et al., 2012) and NOPr (Thompson et al., 2012), confirming the presence of classical

GPCR characteristics. GPCRs are the most abundant class of membrane proteins and act as cellular switches, mediating a vast array of physiological processes. All GPCRs are characterised by 7 membrane spanning helices connected by 3 intracellular loops alternating with 3 extracellular loops, as well as an extracellular N-terminus containing multiple N-glycosylation sites, and an intracellular carboxyl tail containing phosphorylation sites (Law et al., 2000; Rana et al., 2001). GPCRs signal via heterotrimeric guanine nucleotide binding proteins or G-proteins, which associate with the intracellular face of the receptor. G-proteins are composed of 3 subunits, Gα, Gβ and Gγ.

When the receptor is in an inactive state, Gα is bound to GDP and is tightly associated with the obligate heterodimer Gβγ. Ligand binding to the extracellular surface or transmembrane domains of GPCRs induces a conformational change in the Gα subunit, facilitating exchange of GDP for GTP and causing Gα and Gβγ to dissociate. Both Gα and

Gβγ are then able to couple to distinct cellular signalling pathways and effector molecules

(McCudden et al., 2005).

7 1.3 MOPr Structure and G-Protein Signalling

The recent crystallisation of mouse MOPr bound to the selective irreversible antagonist β- funaltrexamine (β-FNA) has provided a number of insights into important structural features of the receptor (Manglik et al., 2012). Compared with other GPCRs, the ligand- binding pocket of MOPr is relatively exposed to the extracellular surface, which likely contributes to the rapid dissociation kinetics observed for many opioid ligands. The binding pocket is made up of elements of transmembrane (TM) domains and extracellular loops (ECLs), including TM3, TM5, TM6 and TM7, and possibly residues from ECL2

(Manglik et al., 2012; Serohijos et al.¸ 2011). Coupling of the receptor to associated G- proteins and effector molecules is mediated by intracellular loops (ICLs), including ICL2 and ICL3, as well as by the C-terminal region (Gether, 2000). The intracellular regions, in particular the C-terminal tail, also contain a number of important phosphorylation sites that regulate receptor desensitisation, internalisation and resensitisation (Williams et al., 2013).

MOPr preferentially couples to the pertussis toxin (PTX)-sensitive Gαi/o family of G- proteins, as well as the closely related PTX-insensitive Gαz and Gα16 (Chakrabarti et al.,

1995; Connor & Christie, 1999; Garzon et al., 1997; Laugwitz et al., 1993). One of the hallmarks of MOPr signalling is the inhibition of adenylyl cyclase (AC) via Gαi/o subunits.

Co-stimulation of associated Gβγ subunits activates G protein-coupled inwardly rectifying potassium channels (GIRKs) and inhibits voltage-gated calcium channels (ICa). Other

MOPr signalling pathways include phosphorylation of mitogen-activated kinases

(MAPKs), activation of phospholipase C (PLC), and the release of calcium from intracellular stores (Connor & Henderson, 1996; Law et al., 2000). MOPr also signals via

G-protein independent pathways, examples of which include beta-arrestin (β-arr)-mediated extracellular-regulated protein kinase 1/2 (ERK1/2) phosphorylation (Zheng et al., 2008),

8 signal transducer and activator of transcription 5 (STAT5) phosphorylation (Mazarakou &

Georgoussi, 2005), and Src-mediated Ras/Raf-1 recruitment (Zhang et al., 2013).

1.4 MOPr Effectors

1.4.1 MOPr inhibition of adenylyl cyclase

One of the major effectors coupled to MOPr is AC. Activation of MOPr results in the inhibition of AC and subsequent downregulation of cAMP, an important second messenger

(Connor & Christie, 1999). MOPr signalling via AC is highly complex and is AC isozyme specific (Avidor-Reiss et al., 1997). Ten isoforms of AC have been identified, which are differentially regulated by various Gα proteins, Gβγ proteins and other regulatory molecules such as Ca2+/calmodulin (Sadana & Dessauer, 2009). Acute morphine treatment inhibits Gαs-stimulated AC5 and AC6 activity and potentiates Gαs-stimulated AC7 acitivty

(Avidor-Reiss et al., 1997; Kim et al., 2006; Yoshimura et al., 1996). MOPr can also affect

AC activity by alternate pathways. Basal AC2, 4 and 7 activity can be upregulated by PKC phosphorylation, and basal AC1 and 8 activity is stimulated by Ca2+/calmodulin. MOPr is capable of modulating both PKC and intracellular Ca2+, both of which were involved in the opioid stimulation of cAMP in SK-N-SH cells (Sarne et al., 1998). Opioid regulation of

AC isoforms differs between acute and chronic exposure (Nevo et al., 2000; Schallmack et al., 2006). Upregulation of AC activity following chronic opioid exposure, termed

“superactivation”, contributes to the development of opioid tolerance and withdrawal symptoms (Christie, 2008), and is also isozyme specific (Avidor-Reiss et al., 1997).

AC catalyses the formation of cyclic adenosine monophosphate (cAMP), an important second messenger involved in a multitude of cellular processes including regulation of ion channels, release of and gene transcription (Skalhegg & Tasken, 1997).

One of the major actions of cAMP is the activation of cAMP-dependent protein kinase A

9 (PKA). PKA phosphorylates several ion channels important in nociception, including transient receptor potential cation channel, subfamily V, member 1 (TRPV1), N‑methyl‑d

‑aspartate (NMDA) receptor 1 (NR1) and glycine receptor α3 (GLR3, Hucho & Levine,

2007). Pronociceptive GPCRs often cause elevation of cAMP levels (Basbaum et al.,

2009; Hucho & Levine, 2007) and elevated cAMP levels increase neuronal excitability in animal models of acute and chronic pain (Cunha et al., 1999).

cAMP can also signal independently of PKA via exchange signals activated by cAMP

(EPACs). EPACs are involved in many of the physiological actions of cAMP, including maintenance of circadian rhythms, memory, wound healing and nerve regeneration

(Borland et al., 2009). EPACs have been shown to contribute to inflammatory pain by mediating cAMP to PKC signalling, which can in turn regulate ion channels such as

TRPV1 (Griffin et al., 2005; Hucho et al., 2005). Although the effects of MOPr-mediated inhibition of EPACs via AC inhibition has not yet been described, EPAC mRNA and protein expression in dorsal root ganglia are increased in mouse models of neuropathic pain (Eijkelkamp et al., 2013), and decreasing EPAC levels in vivo has been shown to prevent the development of chronic hyperalgesia and inflammatory pain in transgenic mice

(Wang et al., 2013). Another PKA-independent consequence of acute AC inhibition is the modulation of voltage-dependent ion channels (Ih, Ingram & Williams, 1994; Svoboda &

Lupica, 1998). Ih is a non-selective cation channel that is activated at hyperpolarized membrane potentials, causing an inward current that depolarizes the membrane back towards threshold. Elevated cAMP causes Ih to be activated at less negative potentials, and thus the decrease in cAMP following MOPr activation reduces activation of Ih channels, reducing neuronal excitability (Ingram & Williams, 1994).

10 1.4.2 MOPr activation of potassium channels

MOPr activation of potassium channels causes a net efflux of intracellular K+ ions, resulting in membrane hyperpolarisation and subsequent inhibition of neuronal firing (Law et al., 2000). MOPr-mediated hyperpolarisation of neurons was first demonstrated in guinea pig locus coeruleus neurons in 1980, although at that point the mechanism had not been defined (Pepper & Henderson). It has since been demonstrated that MOPr can

+ activate several types of potassium channels, GIRK (Kir3 channels), voltage-gated K (Kv)

+ channels, and possibly ATP-sensitive K (KATP) channels, although direct evidence for this last interaction is lacking. GIRK channels are tetramers composed of GIRK1 – 4 subunits

(Hibino et al. 2010). Upon stimulation of MOPr, GIRKs are activated via coupling of a

Gβγ subunit to each of the GIRK subunits, resulting in efflux of intracellular K+ and membrane hyperpolarisation, ultimately reducing neuronal excitability (Corey & Clapham,

2001). MOPr couples to GIRK1/2 channels expressed post-synaptically in the spinal cord

(Marker et al., 2005). Studies in GIRK1/2 knockout mice show a significant reduction in morphine-induced spinal antinociception (Ikeda et al., 2000; Marker et al, 2002; Mitrovic et al., 2003; Marker et al., 2004), and met-enkephalin-induced hyperpolarisation of locus ceruleus neurons (Torrecilla et al., 2002). DAMGO has also been shown to activate GIRKs pre- and post-synaptically in GABAergic neurons in rat periaqueductal grey (PAG,

Vaughan et al., 2003).

MOPr activation of Kv channels causes inhibition of GABA-mediated neurotransmission, resulting in disinhibition of descending antinociceptive pathways (Vaughan et al., 1997).

The activation of Kv channels by MOPr was first demonstrated in non-pyramidal hippocampal neurons in rats (Wimpey & Chavkin, 1991), and has also been demonstrated in rat PAG (Vaughan & Christie, 1997) and basolateral neurons (Finnegan et al.,

2006). The identification of specific Kv channels activated by MOPr has been addressed

11 using knockout mice, as Kv channel blockers such as 4-aminopyridine (4-AP) and tetraethylammonium (TEA) are non-selective. Mice with the Kv1.1 gene inactivated showed lower morphine-induced antinociception compared with wild-type mice (Galleotti et al., 1997; Clark & Tempel, 1998), indicating that activation of Kv1.1 channels plays a role in MOPr-mediated analgesia.

+ Another type of inwardly rectifying K channel regulated by GPCRs is the KATP channel.

Both Gi/o α and βγ subunits are able to open KATP channels (Sanchez et al., 1998; Wada et al., 2000). KATP channel blockers such as glibenclamide and gliquidone antagonise the antinociceptive effects of morphine in rodent spinal cord, brain, and peripheral neurons, whereas KATP channel openers such as diazoxide potentiate the analgesic effects of morphine (Ocana et al., 2004) and fentanyl (Rodrigues et al., 2005), however there is no direct evidence of MOPr activation of these channels.

1.4.3 MOPr inhibition of calcium channels

One of the inhibitory actions of opioids is the suppression of release by inhibiting presynaptic Ca2+ channels and subsequent neuronal firing in the peripheral and central nervous systems. MOPr couples primarily to voltage-gated Ca2+ channels,

2+ including N-, L- and P/Q-type Ca channels (Law et al., 2000). GPCR inhibition of ICa channels is predominantly mediated by the direct coupling of Gβγ subunits to the channel, physically altering the conformation of the channel and inhibiting Ca2+ ion influx

(Dolphin, 2003). ICa are composed of a considerable number of possible combinations of channel subunits, which may be differentially regulated by GPCRs (Randall, 1998).

Furthermore, there are multiple subtypes of both Gβ and Gγ proteins, resulting in a vast number of possible GPCR-channel complexes (Tedford & Zamponi, 2006). Adding to the complexity of GPCR modulation of ICa channels is the ability of both Gα and Gβγ to

12 regulate the channels via second messenger systems such as PLC and cAMP, which are modulated by MOPr (Diverse-Pierluissi et al., 2000), and by G protein independent mechanisms (Iegorova et al., 2010). The functional coupling of MOPr to N-type and P/Q- type ICa has been observed in rat spinal DRG (Schroeder & McCleskey, 1993; Rusin &

Moises, 1995; Schroeder et al., 1991), as well as a number of brain regions including rat

LC (Connor et al., 1999a; Ingram et al., 1997), mouse PAG (Connor et al., 1999b; Kim et al., 1997), and in heterologous expression systems (Seward et al., 1991; Borgland et al.,

2003; Morikawa et al., 1995).

1.4.4 MOPr release of intracellular calcium

2+ 2+ Alteration of the intracellular concentration of Ca ([Ca ]i) is a key regulator of many

2+ cellular processes. In addition to preventing the elevation of [Ca ]i by inhibiting ICa,

2+ 2+ MOPr activation can also elevate [Ca ]i by mobilising Ca from intracellular stores in the endoplasmic reticulum (ER, Samways & Henderson, 2006). In some expression systems

MOPr activation alone can stimulate Ca2+ release from ER (Harrison et al., 1999; Hauser et al., 1996; Zimprich et al., 1995), but in many other instances MOPr potentiates the Ca2+ mobilisation stimulated by co-activation of Gq coupled receptors (Connor & Henderson,

1996; Okajima et al., 1993). The mechanism for this receptor cross-talk is not fully understood, but is likely to involve MOPr mediated modulation of PLC and inositol (1,4,5) triphosphate (IP3, Werry et al., 2003; Zimprich et al., 1995).

Activation of MOPr usually results in a suppression of neurotransmitter release, due to inhibition of Ca2+ cellular influx (Christie et al., 2000). On the other hand, the elevation of

2+ 2+ [Ca ]i resulting from intracellular Ca mobilisation may be sufficient to stimulate neurotransmitter release (Xu & Gintzler, 1992). Other studies have shown that elevation

2+ of [Ca ]i is important in mediating morphine induced analgesia. PLCβ3 knockout mice

13 show a reduced response to morphine (Xie et al., 1999), and when various components of the IP3 pathway are blocked or inhibited, mice also show a reduced sensitivity to opioid

2+ induced analgesia (Aoki et al., 2003; Narita et al., 2000). MOPr elevation of [Ca ]i can also activate MAP kinases, which are involved in a wide range of cellular processes (Law et al., 2000).

1.4.5 MOPr and the MAP kinase cascade

As well as the regulation of cellular signalling and neurotransmission, in recent years it has become apparent that opioid receptors are intimately involved with many other cellular signalling processes such as cell proliferation, differentiation and apoptosis. In many cases,

GPCR control of these processes is via activation of mitogen-activated kinases (MAPKs;

Luttrell, 2002). The 3 major classes of mammalian MAPKs are the extracellular-signal- regulated kinases (ERKs), the Jun N-terminal kinases (JNKs) and the p38 kinases

(p38MAPKs; Law et al., 2000). All 3 classes can be stimulated, i.e. phosphorylated following activation of Gi G-proteins via multiple signalling pathways involving both Gα and Gβγ-coupled mechanisms. Gαi can activate ERK1/2 via Gαi-mediated inhibition of

AC, which decreases the inhibitory effect of PKA on c-Raf. The Ras-c-Raf signalling module is then able to activate ERK1/2 via phosphorylation by MEK1/2 (Goldsmith &

Dhanasekaran, 2007). Gβγ has also been shown to stimulate ERK1/2 directly via Ras- dependent pathways, although the mechanism by which Gβγ couples to Ras varies between receptors, cell types and assay conditions (Luttrell, 2002). ERK1/2 can be activated via G- protein independent mechanisms involving β-arrestin (Macey et al., 2006; Shenoy &

Lefkowitz, 2005; Zheng et al., 2008), and G-protein receptor kinases (GRKs, Macey et al.,

2006). MOPr stimulation of ERK1/2 has been observed in a number of heterologous expression systems via various signalling mechanisms, including activation of phospholipase C (PLC), phosphatidylinositol-3 kinase (PI3K), epidermal growth factor

14 receptor (EGFR) transactivation, and intracellular Ca2+ release (Law et al., 2000). The ability of MOPr to activate ERK1/2 varies with cell type (Mouledous et al., 2004;

Trapaidze et al., 2000), and ERK1/2 regulation also differs with acute and chronic MOPr activation (Bilecki et al., 2005; Mouledous et al. 2004). ERK1/2 signalling is intimately involved in the long-term cellular adaptations to chronic MOPr stimulation, although whether MOR activates these pathways directly is unclear (Christie, 2008; Polakiewicz et al., 1998; Tan et al., 2003b).

There is some evidence for the involvement of JNKs and p38MAPKs in MOPr signalling, with some studies showing JNKs are activated in response to MOPr stimulation (Kam et al., 2004; Melief et al., 2010). Another study reported increased MOPr expression via p38MAPks when JNKs were inhibited (Wagley et al., 2013). JNKs have also been implicated in the development of MOP tolerance (Melief et al., 2010), but not desensitisation (Levitt & Williams, 2012).

1.4.6 Other MOPr effectors

The signalling cascades activated by MOPr are complex, and involve many additional effector molecules including protein kinase B (Akt). Akt mediates cellular processes such as growth, survival, and glucose uptake, and in the case of opioid signalling, is thought to play a neuroprotective role (Muller et al., 2004). MOPr activation has been shown to cause an increase of phosphorylated Akt (pAkt) via PI3K mediated pathways in CHO cells and in a brain-region specific manner in rat brain. More recently, several novel signalling pathways have been identified for MOPr, including STAT5 phosphorylation (Mazarouki &

Georgoussi, 2005) and Src-mediated Ras/Raf-1 recruitment (Zhang et al, 2013) in heterologous expression systems. In both of these studies, activated MOPr was converted into a receptor tyrosine kinase-like entity. Ligand-binding to MOPr caused

15 phosphorylation of the conserved AA motif at residues 336-339. The phosphorylated motif was shown to be a docking site for the direct binding and phosphorylation of STAT5, which regulates gene transcription (Mazarouki & Georgoussi, 2005). Phosphorylated

Tyr336 also served as a docking site for the small regulatory molecule son-of-sevenless

(Zhang 2013). MOPr-mediated phosphorylation son-of-sevenless caused recruitment and activation of Ras/Raf1, which in turn was shown to upregulate AC activity (Zhang 2013).

1.4.7 MOPr desensitisation and endocytosis

Acute MOPr stimulation causes rapid desensitisation and internalisation of the receptor, and chronic MOPr activation results in long-term cellular adaptations that ultimately lead to opioid tolerance (Christie, 2008). The development of desensitisation and tolerance following receptor activation is a characteristic of GPCRs. A generally accepted mechanism for the rapid desensitisation of GPCRs is receptor phosphorylation by G protein-coupled receptor kinases (GRKs) and subsequent β-arrestin binding. In this model,

GRKs are recruited by Gβγ and phosphorylate agonist occupied GPCRs, causing recruitment and binding of β-arrestins to the phosphorylated receptor, uncoupling it from associated G-proteins and thus preventing further signalling (Kelly et al., 2008). The receptor/arrestin complex is then trafficked to clathrin-coated pits and internalised, whereupon it is either recycled to the plasma membrane or else targeted to lysosomes for degradation (Qiu et al., 2003). Receptor endocytosis is thought to contribute directly to

MOPr desensitisation by reducing the number of available receptors at the plasma membrane, yet morphine, which is well known to produce tolerance, does not appear to induce significant levels of MOR endocytosis, suggesting additional pathways for the development of receptor desensitisation and tolerance (Alvarez et al., 2002).

16 The third intracellular loop and the C-terminal tail of GPCRs are important domains for the regulation of receptor phosphorylation, desensitisation and internalisation (Lefkowitz,

1998). Hierarchical, multi-site phosphorylation of the C-terminal domain is required for receptor endocytosis (El Kouhen et al., 2001). While morphine induces phosphorylation at the S375 residue of the C-terminal domain, DAMGO and other agonists that can efficiently stimulate receptor endocytosis cause subsequent phosphorylation of the T370 residue, which in turn results in endocytosis (Doll et al., 2011, Grecksch et al., 2011). The mechanism for receptor desensitisation also appears to be ligand-dependent. DAMGO has been shown to induce MOPr desensitisation via the GRK/β-arrestin dependent pathway, whereas morphine does not, except in the case of GRK2 overexpression (Zhang et al.,

1998). Morphine seems able to induce desensitisation in the absence of β-arrestin, via pathways involving PKC (Bailey et al., 2009; Chu et al., 2008; Chu et al., 2010; Johnson et al., 2006). There is some evidence to suggest that the desensitisation pathways activated are directly related to efficacy of the ligand at the receptor, that is, morphine is unable to activate a sufficient number of Gβγ subunits to recruit GRKs to the receptor (Connor et al.,

2004; Hull et al., 2010; McPherson et al., 2010). However, the efficacy of opioid ligands in activating G-proteins has been shown to be distinct from their ability to promote MOPr endocytosis (Borgland et al., 2003). Furthermore, there is additional evidence that morphine binding to MOPr results in a specific ligand/receptor conformation that PKC recognizes and is able to bind to (Ingram & Traynor, 2009). Although the mechanisms for activation of specific MOPr endocytotic and desensitisation pathways are not well understood, it is clearly a more complex system than one based solely on ligand efficacy, and may be dependent on specific receptor conformations induced by individual ligands.

17 1.5 MOPr Ligands

1.5.1 Endogenous MOPr ligands

Shortly after the discovery of opioid receptors, the endogenous opioids met- and leu- enkephalin were characterized (Hughes et al., 1975), followed by β-endorphin (Li &

Chung, 1976) and the (Goldstein, 1979). Since then, a number of endogenous opioid have been identified. Most of these originate from one of three distinct precursor proteins; proenkephalin (pEnk), prodynorphin (pDyn), and proopiomelanocortin

(POMC, Akil et al., 1984). Endogenous opioids derived from these precursors bind to

MOPr, DOPr and KOPr with varying affinities, however have no affinity for NOPr

(Pawson et al., 2014). pEnk gives rise to met- and leu-enkephalin, which show highest affinity for DOPr, lower affinity for MOPr, and little affinity for KOPr. POMC encodes β- endorphin as well as other non-opioid peptides like adrenocorticotropic hormone (ACTH;

Akil et al., 1984). β-endorphin couples to MOPr and DOPr, with low affinity for KOPr. pDyn contains coding for the dynorphins, which preferentially bind to KOPr, but also couple to MOPr and DOPr (Pawson et al., 2014). The highly selective MOPr agonists endomorphin-1 (EM1) and endomorphin-2 (EM2) have been isolated from various tissues

(Wang et al., 2002b; Zadina et al., 1997), however the coding gene and precursor protein have yet to be identified. It is thus unclear as to whether EM1 and EM2 are produced endogenously. The principle endogenous ligand for NOPr is nociceptin or orphanin FQ

(N/OFQ), which is derived from prepronociceptin. N/OFQ does not act at any of the classical opioid receptors, nor is it antagonized by naloxone (Pawson et al., 2014).

1.5.2 Exogenous MOPr ligands

The archetypal MOPr alkaloid agonist is morphine. Morphine has approximately 50 times higher affinity for MOPr than DOPr, and has been shown to have both full and partial agonist activity, depending on assay conditions (Pawson et al., 2014). The morphine

18 metabolite, morphine-6-gluconuride (M6G) is biologically active (Yoshimura et al., 1973).

M6G has a similar affinity to morphine for MOPr, but has been shown to have a higher agonist efficacy than morphine (Osborne et al., 2000). A number of structurally related, semisynthetic morphine derivatives including buprenorphine and oxycodone have been developed, as well as fully synthetic morphine analogues such as methadone. Non- morphinan synthetic alkaloids include the highly efficacious anilidopiperidine fentanyl

(Pawson et al., 2014). D-Ala2-MePhe4-Glyol5-enkephalin (DAMGO) is a synthetic that was designed to be a biologically stable analogue of the DOPr-preferring , but has a high affinity and selectivity for MOPr (Dhawan et al., 1996). MOPr antagonists include naloxone, which has highest affinity for MOPr, but also binds to DOPr and KOPr, and , which has little selectivity for MOPr. The MOPr antagonists with highest selectivity are cyclic peptides related to somatostatin such as CTOP and

CTAP (Dhawan et al., 1996).

1.5.3 Ligand affinity and efficacy

The action of ligands at a receptor can be measured by two separate properties: the ability of the ligand to bind to the receptor, termed ligand affinity, and the ability of the ligand to elicit a physiological effect at the receptor, termed ligand efficacy. Ligand affinity is commonly measured using radioligand-binding assays to determine the concentration of ligand that occupies 50% of the receptor at equilibrium. Ligand “efficacy” is generally measured using concentration-response curves (CRCs) for a particular effector pathway, e.g. cAMP accumulation. However, there is some confusion over the use of the term

“efficacy”. Technically, ligand efficacy refers to the ability of the ligand to activate the receptor and elicit a cellular response (Kelly, 2013). The maximum response that a ligand can elicit is tissue-dependent due to a range of factors including differing levels of g proteins or other effector molecules. Thus, measuring the maximum effect of a ligand at a

19 specific signalling pathways is actually a measure its “intrinsic activity” rather than its efficacy, as two ligands may have a similar maximum response in a particular system or tissue, with very different values of efficacy. Nevertheless, the terms “efficacy” and

“intrinsic activity” are both widely and interchangeably used. Throughout this thesis, the term “efficacy” has been used rather than intrinsic activity to describe the maximum effect of a ligand for each particular system. Full agonists are ligands that elicit the maximum response possible for the system. Partial agonists will cause a sub-maximal response of the system even when all receptors are occupied by ligand. Antagonists are ligands that have affinity for a receptor without causing an observable effect, and inverse agonists bind to receptor and inhibit its constitutive activity (Costa & Herz, 1989) The efficacy of a ligand is tissue-dependent, as receptor expression and the available pool of G-proteins and effector molecules can vary between cell types (Atwood et al., 2011). Furthermore, it is now apparent that ligands vary in their ability to stimulate specific signalling pathways, and thus ligand efficacy is also pathway dependent (Kenakin & Miller, 2010).

1.6 Ligand-biased signalling at MOPr

The early conception of GPCR signalling was that these receptors acted as simple on/off switches with ligand efficacy essentially linear, i.e. after receptor activation, an agonist would have a similar efficacy across the full spectrum of GPCR effectors (Karlin, 1967;

Samama et al., 1993). It was thought that GPCRs oscillated between a single active and inactive state, and that ligands only differed in their ability to stabilise the active state.

Differences in the pharmacological profile of various drugs were attributed to differences in tissue characteristics, as well as the specific kinetics and metabolism of the drug.

Attempts to discover new GPCR drugs with improved pharmacological profiles were focused on designing highly selective compounds for receptor subtypes with distinct anatomical distribution or pharmacological properties. It is now known that GPCRs

20 continually oscillate through a range of active and inactive states, each of which may be differentially stabilized by receptor agonists and antagonists (Kenakin & Miller, 2010).

Receptor ligands can bind with high or low affinity to particular conformations of the receptor, and the resulting ligand/receptor complexes may vary in their ability to couple to different G-protein subtypes and other effectors such as β-arrestin. This in turn can result in preferential activation of a subset of receptor signalling pathways (Ghanouni et al.,

2001). There is a growing body of evidence demonstrating ligand-directed activation of a range of GPCRs (Evans et al., 2011; Rajagopal et al., 2013), including MOPr (DeWire et al., 2013; Pineyro et al., 2007). Plasmon-waveguide resonance (PWR) spectroscopy has demonstrated that DOPr agonists, antagonists and inverse agonists stabilise the receptor in different conformations, furthermore, that that the conformations stabilised are ligand- specific (Alves et al., 2003; Salamon et al., 2002). MOPr agonists have been shown to differ in their ability to stimulate various G-protein subtypes in HEK293 cells. Morphine and buprenorphine activated GαoA to a greater extent than Gαi1, whereas DAMGO, fentanyl, met-enkephalin, leucine-enkaphalin and methadone activated Gαi1 to the same or greater extent than GαoA. The affinities of the agonists for the G-protein subtypes also differed, with DAMGO, E2 and morphine showing significantly greater potencies at GαoA, and E1 showing greater affinity for Gαi1 (Saidak et al., 2006). Agonist selectivity for G- protein activation has been demonstrated in vivo, with some opioid drugs producing different levels of analgesia in Gαi1 and GαoA knockdown mice (Sanchez-Blanquez et al.,

1999). A topic of particular interest at present is the identification of ligands biased towards either G-protein coupled pathways or G-protein independent pathways such as β- arrestin recruitment (Violin & Lefkowitz, 2007). Studies in mice injected with antisense

Gαi oligodeoxynucleotides showed an attenuation of the antinociceptive effects of morphine (Raffa et al., 1994; Sanchez-Blanquez et al., 1993; Sanchez-Blanquez et al.,

1995), whereas β-arrestin KO mice show a reduction in the adverse effects associated with

21 morphine (Raehal & Bohn, 2011; Raehal et al., 2005). This has led to the hypothesis that the desirable characteristics of opioids such as analgesia are mediated by G protein- coupled pathways, while the adverse effects such as tolerance and dependence are mediated via β-arrestin coupled pathways. As such, current efforts to design improved opioid drugs are often focused on developing ligands with significant bias towards G- protein coupled pathways (DeWire et al., 2013). Although the connection between specific ligand/receptor conformations, differential G-protein and β-arrestin activation, and altered signalling has not yet been demonstrated conclusively for MOPr, it is a reasonable supposition that the different effects elicited by various MOPr ligands are due at least in part to differential activation of MOPr effectors.

1.7 OPRM1 Polymorphisms

A major limitation of opioid prescription for the treatment of pain is the highly variable response observed between individuals, both in the analgesic and adverse effects (Skorpen et al., 2008). The risk associated with potentially serious side-effects such as respiratory depression can limit opioid dosing resulting in inadequate pain relief for many individuals

(Boom et al., 2012). Furthermore, some patients will respond to certain opioid drugs and not others, for example many patients who do not respond to methadone treatment for opioid dependence may respond to buprenorphine (Gerra et al., 2014). Several parameters contribute to individual response to opioids, such as drug absorption, distribution and metabolism, as well as the intrinsic efficacy of the drug at the receptor. Genetic factors are likely to play a role in variability of all of these parameters. A number of polymorphisms in the MOPr gene, OPRM1, occur naturally within the population. OPRM1 single nucleotide polymorphisms (SNPs) have been identified in the promoter, intron and coding regions, some of which result in an amino acid change and are associated with functional consequences in vivo or in vitro (Lotsch & Geisslinger, 2005). One SNP of particular

22 interest clinically is the N40D variant, which occurs at relatively high frequencies ranging from 10 – 50% in various populations (Mura et al., 2013). This SNP, an A > G nucleotide substitution at position 118 in exon 1 of OPRM1, causes an asparagine to aspartic acid exchange at residue 40 on the extracellular N-terminal domain of MOPr, removing a putative glycosylation site (Singh et al., 1997). This SNP was first identified by Bergen et al., (1997), and in 1998, Bond et al. reported that β-endorphin had a threefold higher affinity for MOPr-D40 expressed in AV-12 cells and a threefold higher potency for GIRK activation in Xenopus oocytes, although subsequent studies have failed to replicate these findings (Befort et al., 2001; Beyer et al., 2004; Kroslak et al., 2007). Since then a large number of studies have examined associations between the N40D variant and a plethora of clinical effects. Carriers of the G118 allele showed an increased requirement for morphine or fentanyl to treat pain (Campa et al., 2008; Klepstad et al., 2004; Reyes-Gibby et al., 2007) and post-operative pain (Chou et al, 2006a; Chou et al., 2006b; Fukuda et al.,

2010; Sia et al., 2008; Tan et al., 2009; Zhang et al., 2010; Zhang et al., 2011), and in some cases the associated adverse effects such as nausea were also reduced (Sia et al.,

2008; Tan et al., 2009; Tsai et al., 2010). Analgesia, pupillary constriction and nausea were reduced in response to M6G in G118 carriers (Lotsch et al., 2002; Romberg et al.,

2005; Skarke et al., 2003). Overall, these data point to a reduced opioid response in G118 carriers, however a recent meta-analysis concluded that the association is weak, applies to homozygous carriers only, and is of little clinical relevance (Walter & Lotsch, 2009).

Several studies have reported an increased susceptibility to abuse and/or dependence in carriers of the G118 allele (Bart et al., 2004; Drakenburg et al., 2006; Kapur et al., 2007; Szeto et al., 2001), while others have found no association or a protective effect of G118 (Bond et al., 1998; Glatt et al., 2007; Shi et al., 2002; Tan et al., 2003a).

Similarly, studies examining association between the G118 allele and the risk of alcohol

23 dependence have reported a positive association (Bart et al., 2005; Miranda et al., 2010;

Nishizawa et al., 2006; Rommelspacher et al., 2001), a negative association (Du & Wan,

2009; Town et al., 1999), and no association (Bergen et al., 1997; Gscheidel et al., 2000;

Kim et al., 2004a; Loh et al. 2004). However, carriers of the G118 allele appear to consistently show an increased response to alcohol as measured by subjective reports of positive affect (Miranda et al., 2010; Ray & Hutchison, 2004; Ray & Hutchison, 2007;

Ray et al., 2010), cue-induced alcohol cravings (Van den Wildenberg et al., 2007; Wiers et al., 2009), and PET scans of striatal dopamine release (Ramchandani et al., 2011). The endogenous opioid system is involved in the rewarding properties of alcohol, and to this end MOPr antagonists have been used as a pharmacological treatment of alcohol dependence (Thorsell, 2013). Carriers of G118 reported significantly decreased alcohol induced euphoria when treated with naltrexone compared with A/A individuals (Setiawan et al., 2011). Meta-analyses of studies examining the effectiveness of naltrexone treatment for alcoholism have suggested carriers of G118 may respond more favourably with lower relapse rates, although these findings are not consistent (Chamorro et al., 2012; Coller et al., 2011; Thorsell, 2013).

The apparent differences in the response of G118 carriers to alcohol may be mediated in part by changes in activation of the hypothalamic-pituitary-adrenal axis (HPA). The endogenous opioid system and the HPA axis are activated in response to alcohol consumption (Gianoulakis, 1998), as well as stress (Drolet et al., 2001). Studies examining the association between N40D and activation of the HPA axis have shown an increased cortisol response to naloxone (Chong et al., 2006; Hernandez-Avila et al., 2007; Wang et al., 2002a), a dampened adrenocorticorticotropin (ACTH) response to stress (Ducat et al.,

2013), and increased basal cortisol levels with no changes in ACTH (Hernandez-Avila et al., 2003). Taken together, these results are suggestive of increased cortisol activity in

24 G118 carriers, possibly contributing to stress-induced alcohol consumption and improved response to naltrexone therapy for alcoholism.

Another less common OPRM1 SNP has been identified within the N-terminal region of

MOPr. The C17T variant, resulting in an alanine to valine amino acid substitution at aa6 on the N-terminal tail, is quite rare in the overall population but can reach allelic frequencies of more than 20% in some Indian and African populations (Crowley et al.,

2003; Kapur et al., 2007). This variant has been associated with the risk of alcohol, cocaine, tobacco but not opioid use in African-American women (Crystal et al., 2012).

Similarly, other studies have demonstrated a trend towards a higher frequency of T17 carriers in substance abusers, however the small sample sizes of the studies, as well as population-dependent nature of the frequency of the SNP makes it difficult to draw any conclusions from these results (Berrettini et al., 1997; Rommelspacher et al., 2001;

Compton et al., 2003; Crowley et al., 2003). There is no clinical information relating to other, less common MOPr polymorphisms, however some functional differences have been observed in receptor signalling (see Chapter 2, Knapman & Connor, 2014).

1.8 MOPr variants and differential signalling

The idea that specific properties of ligands may stabilise receptor conformations that preferentially activate particular signalling pathways can be expanded to include properties of the receptor as well. Genetic polymorphisms causing amino acid changes may cause subtle differences in the structural flexibility of the receptor, resulting in a set of possible receptor conformations that differ to those exhibited by the wild-type receptor, both in the agonist-occupied and unoccupied states. This idea provides a mechanism by which genetic variation between individuals could translate to variation in response to opioid analgesics.

25 Differences in receptor function have been observed in MOPr receptor variants, however much of the evidence is contradictory and inconclusive, possibly due in part to differences in cellular expression systems, the signalling pathway and ligands assayed, as well as the failure of many of studies to take into account receptor reserve. Although no conclusive evidence is available as to the effect of MOPr polymorphisms on individual response to opioid analgesics, these findings suggest that genetic variation in the MOR may contribute to the highly variable response and warrant further investigation. Chapter 1 has described

MOPr function and signalling, particularly in the context of changing conformational states arising from interaction between the receptor and distinct ligands. In the next chapter, my published review in the British Journal of Pharmacology extends this and discusses what has been reported about the functional effects of MOPr SNPs.

26 Chapter 2

Cellular signalling of non-synonymous single nucleotide polymorphisms of the human μ-opioid receptor

Alisa Knapman and Mark Connor

This review article was published online in 2014 in the British Journal of Pharmacology, doi: 10.1111/bph.12644. The manuscript was prepared by myself and Mark Connor.

27

28 Summary

There is significant variability in individual responses to opioid drugs, which is likely to have a significant genetic component. A number of non-synonymous single nucleotide polymorphisms (SNPs) in the coding regions of the µ-opioid receptor (MOPr) gene

(OPRM1) have been postulated to contribute to this variability. Although many studies have investigated the clinical influences of these MOPr variants, the outcomes are reported in the context of thousands of other genes and environmental factors, and we are no closer to being able to predict individual response to opioids based on genotype. Investigation of how MOPr SNPs affect receptor expression, coupling to second messengers, desensitisation and regulation is necessary to understand how subtle changes in receptor structure can impact individual response to opioids. To date, the few functional studies which have investigated the consequences of SNPs on the signalling profile of the MOPr in vitro have shown that the common N40D variant has altered functional responses to some opioids, while other, rarer, variants display altered signalling or agonist-dependent regulation. Here, we review the available data on the effects of MOPr polymorphisms on receptor function, expression and regulation in vitro, and discuss the limitations of the studies to date. Whether or not MOPr SNPs contribute to individual variability in opioid responses remains an open question, in large part because we have relatively little good data about how the amino acid changes affect MOPr function.

Keywords: A118G; pharmacogenomics, analgesia, addiction, tolerance, dependence

29 Abbreviations

aa amino acid

AC adenylyl cyclase

β-CNA β-chlornaltrexamine

CaM calmodulin cAMP cyclic adenosine monophosphate

CaMKII Ca2+/calmodulin-dependent protein kinase II

CRE cAMP response element

DAMGO ([D-Ala2, N-MePhe4, Gly-ol]-enkephalin)

ECL extracellular loop

ELISA enzyme-linked immunosorbent assay

ERK extracellular signal regulated kinase

GIRK G protein gated, inwardly rectifying potassium

channel

GRK G protein coupled receptor kinase hMOPr human µ-opioid receptor

HPA hypothalamic-pituitary-adrenal

ICa voltage gated Ca channels

ICL intracellular loop

M-6-G morphine-6-glucuronide

MAP kinase mitogen activated protein kinase

MOPr µ-opioid receptor mRNA messenger ribonucleic acid

OPRM1 opioid receptor mu 1 pERK phosphorylated ERK1/2

30 PKA protein kinase A

PTX pertussis toxin

SNP single nucleotide polymorphism

STAT5 signal transducer and activator of transcription 5

TM transmembrane

31 Introduction

Opioid analgesics are the most important classes of drug used for the treatment of moderate to severe pain. Opioids elicit powerful analgesic effects, yet they are also associated with a number of adverse effects such as respiratory depression, constipation, nausea and sedation (Moore & McQuay, 2005; Dahan et al., 2010; Noble et al., 2010). The development of tolerance to opioid analgesia, coupled with the associated adverse effects, limits the usefulness of opioid therapy in the treatment of long-term and chronic pain.

Opioid misuse is also a major social problem in many countries (Dhalla et al., 2011).

There is significant variation between individuals in both the analgesic effect of opioid drugs and the degree of adverse effects experienced. The risk of serious adverse events such as respiratory depression can limit dosing with the result that many individuals receive inadequate pain relief (Skorpen & Laugsand, 2008). Furthermore, as tolerance develops over time, the escalating opioid doses that are required to maintain adequate analgesia can cause intolerable side effects (Corbett et al., 2006). There is also an apparently heritable predisposition towards opioid abuse and addiction (Merikangas et al.,

2008). A number of elements may affect final individual response to opioids including drug absorption, distribution and metabolism, as well as the intrinsic efficacy of the drug at the receptor and variation in receptor signalling function, agonist regulation and downstream effector pathways. Genetic factors such as differences in protein sequence, regulatory element function and potentially complex epigenetic regulation of protein expression contribute to variability in all these parameters (Lotsch et al., 2005; Skorpen &

Laugsand, 2008). Understanding these components could result in the ability to better predict clinical outcomes when prescribing opioid analgesics, reducing the number of patients receiving an inappropriate dose of opioid by potentially limiting the development

32 of tolerance and dependence. Rational dosing would also likely increase the number of patients who benefit from opioid therapy.

Clinically important opioid analgesics act by binding to the μ-opioid receptor (MOPr)

(Matthes et al., 1996; Alexander et al.; 2011), making MOPr a prime candidate for contributing to the genetic component of inter-individual differences in opioid response.

The MOPr is a typical Class A G protein coupled receptor (GPCR, Alexander et al.; 2011).

Many single nucleotide polymorphisms (SNPs) within the ORPM1 gene have been identified in , and a number of these are non-synonymous changes in the coding regions, meaning that there is an amino acid (aa) substitution resulting in an alternative receptor isoform (LaForge et al., 2001; Ikeda et al., 2005; Lotsch et al., 2005;

Ravindranathan et al., 2009; Fortin et al., 2010). There are good reasons to consider the potential of MOPr SNPs to contribute to the clinical variability of opioid responses. GPCR signalling is complex, with the notion of simple, linear and robust re-arrangements of protein structure being required for signal transduction no longer accepted. Thus, the possibility that single amino acid substitutions can lead to subtle or profound changes in the way receptors signal is very real, and has been demonstrated for several GPCRs

(Thompson et al., 2008; Zhang et al., 2013a). Further, commonly prescribed opioids such as morphine and buprenorphine have relatively low efficacy, and even modest differences in receptor expression or efficiency of signal transduction could have a significant impact on individual response to these drugs. Finally, clinically used opioids are chemically diverse, and are likely to have subtly different structural features of the MOPr determining their signalling – potentially leading to distinct effects of non-synonymous SNPs on different drugs.

33 An additional level of complexity when considering the functional consequences of SNPs arises from the large number of putative splice variants of MOPr which have been described (Mizoguchi et al., 2012; Pasternak & Pan, 2013). Although the functional role of alternatively spliced OPRM1 transcripts is not yet well established, a single non- synonymous amino acid change could conceivably have distinct effects on different splice variants of the receptor. For the most part, this remains unexplored.

Many studies have examined potential associations between MOPr SNPs and various clinical outcomes, such as the degree of pain relief in response provided by opioids or the prevalence of substance abuse. These clinical reports are often contradictory and there is no clear consensus as to the effect of any polymorphism on disease susceptibility or the outcomes of drug administration. This is presumably in part due to relatively small sample sizes in most studies, as well as a range of confounding influences such as overall genotype and environment (reviewed in Lotsch et al., 2005). Far fewer studies have investigated the molecular consequences of OPRM1 SNPs on receptor function and signalling in vitro, and results from these studies are also conflicting. Nevertheless, in vitro experiments have led to intriguing insights into MOPr function, and in this review we focus on the effects of naturally occurring, non-synonymous SNPs in the coding region of

OPRM1 on MOPr function. The SNPs considered here, the corresponding amino acid exchanges and their position on the MOPr are summarized in Table 1 and Figure 1.

The µ-opioid receptor

The MOPr is a Class A rhodopsin-like GPCR, with a relatively short extracellular N- terminal domain (66 aa), 7-membrane spanning domains and an intracellular carboxy- terminal “tail” (70 aa) that includes a putative “helix 8” domain tethered to the plasma membrane by a palmitoyl residue (Manglik et al., 2012). Opioid ligands are thought to

34 approach the receptor from the extracellular space, engaging with the receptor by interacting with a binding pocket formed by elements of transmembrane (TM) domains

TM3, TM5, TM6 and TM7, and possibly residues in extracellular loop (ECL) 2 (Serohijos et al., 2011; Manglik et al., 2012). G-protein interactions are mediated through intracellular domains, including intracellular loops (ICL) 2 and 3, and the C-terminal region. The intracellular regions of MOPr, particularly the C-terminal domain, also contain important phosphorylation sites regulating receptor desensitisation, internalisation and resensitisation (Williams et al., 2013).

MOPR mutation diagram

Figure 1: Naturally occurring, non-synonymous OPRM1 variants reported, and their position on the MOPr protein. Residues where an amino acid exchange occurs are indicated in red.

35 The MOPr modulates a diverse range of physiological systems, including nociception and analgesia, reward and euphoria, immune function, stress responsivity, respiration and gut motility (Jordan & Devi, 1998; Kreek et al., 2005). The most well characterized signalling pathways of MOPr proceed via activation of heterotrimeric G proteins or β-arrestin (Law et al. 2000). MOPr can couple to a number of different G proteins, including pertussis- toxin (PTX) sensitive Gαi/o subunits, the closely related Gαz, and Gα16 (Connor & Christie,

1999). Canonical coupling of MOPr includes Gαi/o inhibition of adenylyl cyclase (AC),

Gβγ subunit activation of G protein-coupled inwardly rectifying potassium channels

2+ (GIRKs) and inhibition of voltage gated Ca channels (ICa), as well as activation of MAP kinases. Examples of G protein-independent signalling of MOPr include β-arrestin- mediated extracellular signal regulated kinase 1/2 (ERK1/2) activation (Zheng et al, 2008), signal transducer and activator of transcription 5 (STAT5) phosphorylation (Mazarakou &

Georgioussi, 2005) and Src-mediated Ras/Raf-1 recruitment (Zhang et al., 2013b).

The MOPr, like all GPCRs, has many active conformations (Pinyero et al., 2007; Kenakin

& Miller, 2010; Manglik et al., 2012). In their unbound state, GPCRs constantly oscillate through a range of possible conformational states. Ligands bind to GPCRs and stabilize subsets of conformational states, some of which couple to and activate downstream effectors (agonists), while other are not coupled to effectors, and when ligands bind they prevent downstream signalling (antagonists). The stabilisation of subsets of conformations by a ligand may lead to preferential activation of a restricted set of signalling pathways, leading to ligand-specific patterns of signalling and receptor regulation – also known as ligand-biased signalling or functional selectivity. The MOPr binds an array of structurally diverse ligands and interacts with many effector and regulatory proteins providing a fertile system for ligand biased signalling. (Kenakin, 2002; Massotte et al., 2002; Saidak et al.,

2006).

36 The corollary of structurally distinct agonists and effector molecules preferentially coupling via subsets of receptor conformations is that changes in the molecular structure of the receptor itself are likely to affect receptor conformation (Abrol et al., 2013; Cox,

2013). Thus, amino acid changes resulting from SNPs have the potential to affect MOPr signalling globally or in a ligand-dependent manner by affecting the ability of a ligand to bind to the receptor, altering the conformation of the ligand-receptor complex and/or affecting the ability of this complex to couple to G-proteins and associated signalling or regulatory pathways.

Functional studies of MOPr SNPs

Most functional studies of human MOPr (hMOPr) use heterologously expressed receptors in an immortalized cell line such as CHO-K1, HEK-293 or AtT-20. The physiological relevance of subtle differences in signalling exhibited by MOPr variants in these highly engineered expression systems is difficult to predict, and making direct comparisons between receptor signalling profiles in different expression systems may be problematic as different cell lines vary in the available pool of G-proteins, effector molecules and regulatory proteins (e.g. Atwood et al., 2012). Nevertheless, µ-opioid receptors are naturally expressed in a wide variety of cell types, and there is unlikely to be “one true path” for receptor activation and regulation. Thus, studies in diverse systems are probably necessary to capture the possible consequences of variations in receptor structure.

However, in order to make comparisons between polymorphic variants meaningful, careful attention needs to be paid to receptor expression levels and the nature of the signalling assays (Connor et al., 2004). While there seem to be functional differences between MOPr

SNPs and the most common form of the receptor, many variants have been superficially described and making firm conclusions about the consequences of variations in MOPr sequence is limited by the experimental conditions used to study them.

37 Table 1: Summary of non-synonymous MOPr variants in the protein coding region, their corresponding OPRM1 SNP, exon and MOPr protein domain. See references in the text for original reports.

AA Exchange MOPr Domain SNP Exon N40D N-terminus 118 A > G 1 A6V N-terminus 17 C > T 1 S42C N-terminus 124 T > A 1 D51N N-terminus 151 G > A 1 G63V N-terminus 188 G > T 1 S66F N-terminus 197 C > T 1 L85I TM1 253 C > A 1 S147C TM3 440 C > G 2 N152D TM3 454 C > G 2 R181C ICL2 541 C > T 2 N190K ICL2 570 A > T 2 C192F TM4 575 G > T 2 R260H ICL3 779 G > A 3 R265H ICL3 794 G > A 3 S268P ICL3 802 T > C 3 D274N ICL3 820 G > A 3 V293I TM6 877 G > A 3

38 N-terminal domain SNPs

N40D

The N40D variant is the most commonly occurring OPRM1 SNP, with an allelic frequency ranging from 10 - 50% within various populations (Mura et al., 2013). The N40D SNP is in the N-terminal extracellular domain of MOPr, and removes one of 5 putative asparagine-linked glycosylation sites in this region (Table 1, Singh et al., 1997). First reported by Bergen et al. (1997), many studies have examined associations between the

D40 allele and physiological and clinical parameters including nociception, altered response to opioid analgesics, opioid and alcohol dependence and hypothalamic-pituitary- adrenal (HPA) axis responses (Kreek et al, 2005; Walter and Lotsch, 2009). Most of the association studies report that carriers of the D40 allele have a reduced response to opioids, although some studies have reported the opposite, and others no effect at all (reviewed in

Diatchenko et al. 2011). A recent meta-analysis of the clinical effects of the N40D variant in pain management concluded that knowing a patient’s allele(s) at position 118 in OPRM1 would have little impact on treatment (Walter & Losch, 2009), although the number of studies available for review was small. The D40 allele also has been associated with an increased, decreased or unchanged susceptibility to drug use and dependence, (reviewed in

Mague et al., 2010).

Regulation of N40D expression

Regardless of any impact on the function of the µ-opioid receptor, the possibility that the nucleotide or amino acid substitutions may affect MOPr expression levels needs to be considered. There is some evidence for reduced MOPr expression associated with the

G118 allele (or its murine ortholog). It was reported that in cortex and from the of A118G heterozygotes, there was significantly less G118 mRNA (1.5 - 2.5-fold) than A118 mRNA (Zhang et al., 2005a). A similar reduction in mRNA was found in a

39 knock-in mouse model with an orthologous A112 to G112 mutation (Mague et al., 2009).

A potential explanation for the reduced levels of G118 mRNA was provided by Oertel et al., (2012), who deduced that the G118 allele has an extra methylation site introduced by the guanine nucleotide, which was suggested to inhibit compensatory upregulation of

MOPr mRNA in chronic opioid users. It is possible that this epigenetic regulation also results in lower levels of G118 mRNA in basal conditions.

The loss of the potential glycosylation site, N40, may also contribute to lower levels for the D40 allele, although this has not been consistently reported (Zhang et al., 2005a; Oertel et al., 2009). In mice, the molecular mass of MOPr in 112G/G animals (55 kDa) is lower than 112A/A mice (62 kDa), whereas the molecular mass of deglycosylated MOPr was identical (42 kDa) for both variants, indicating less glycosylation in 112G/G mice (Huang et al., 2012). The G/G mice also have lower MOPr expression compared with A/A mice (Mague et al., 2009; Wang et al., 2012). Findings of lower expression extend to cells lines (Zhang et al., 2005a; Huang et al., 2012), with a shorter half life of D40 (12 hr) compared with N40 (28 hr) in CHO cells. Enzymatic deglycosylation of MOPr also decreased receptor expression in HEK-293 cells by 90%

(Kroslak et al., 2007). Decreased mRNA stability, potential epigenetic repression and incomplete glycosylation could all contribute to reduced D40 receptor expression, potentially providing a mechanism for greater opioids requirement in D40 carriers (Mura et al., 2013).

Second Messenger Coupling

The consequences of the N40D substitution on the signalling profile of MOPr are not well understood and despite the intense research into the clinical effects of the D40 variant, only a handful of functional studies in cells have been performed on this variant (Table 2). The

40 first reported functional consequences of a MOPr SNP was a three-fold increase in the affinity of β-endorphin for the D40 variant and a three-fold increase in the potency of β- endorphin to activate GIRK channels co-expressed D40 in Xenopus laevis oocytes (Bond et al., 1998). No differences in binding or signalling were reported for other opioids including DAMGO, endomorphin 1 and 2, and enkephalins. Unfortunately, these provocative results were based on very limited quantification of the cellular responses to activation of the N40 and D40 alleles, and no statistical analysis was included. Subsequent studies looking at different signalling pathways in other expression systems have failed to find differences in β-endorphin potency at N40 and D40 (Befort et al., 2001; Beyer et al.;

2004; Kroslak et al., 2007).

N40 inhibition of AC has been examined in several studies in HEK 293 cells (Beyer et al.

2004; Kroslak et al., 2007; Fortin et al., 2010). Unfortunately, these studies did not use

N40 and D40 cell lines with equivalent receptor expression, and receptor reserve for AC inhibition was not assessed. Beyer and colleagues (2004) found no differences in the effects of morphine, morphine-6-glucuronide (M-6-G) or β-endorphin to inhibit acute cyclic adenosine monophosphate (cAMP) accumulation despite 7 fold lower expression of

D40 than N40 in their cells. Fortin et al. (2010) also found no difference in how DAMGO, endomorphin-1 or leu-enkephalin modified cAMP-dependent gene transcription in cells acutely transfected with D40 and N40 constructs. This strategy produced apparently equivalent levels of receptor expression (measured by enzyme-linked immunosorbent assay [ELISA]) but the absolute levels were not reported. By contrast, Kroslak et al.

(2007) reported a decreased potency of morphine, methadone and DAMGO but not β- endorphin to inhibit cAMP accumulation in cells expressing D40, however, this was associated with a 66% lower expression of D40 compared with N40. It is difficult to explain the differences between these studies, particularly in the absence of information

41 about relative efficacy. Studies using cAMP-dependent gene expression assays to measure

MOPr activity have prolonged incubation with agonist and the response measured reflects the integrated outcome of acute inhibition of AC as well as agonist-dependent uncoupling, internalisation and possible recycling or degradation of the receptor, any of which could be altered by the D40 polymorphism (Connor et al., 2004; Dang & Christie 2012). Likewise, possible differences in the acute regulation of D40 and N40 variants of the MOPr over the time course of acute cAMP accumulation assays could also confound their outcomes

(Connor et al., 2004).

The activity of N40 and D40 have also been compared by measuring inhibition of ICa in acutely transfected sympathetic neurons (Margas et al., 2007), HEK293 cells (Lopez Soto

& Raingo, 2012) as well as mice “humanized” with A118 and G118 knock-in (Mahmoud et al., 2011; Ramchandari et al., 2011). Opioid receptor inhibition of ICa is via direct Gβγ- subunit inhibition of channel gating. In both HEK293 cells and sympathetic neurons,

DAMGO inhibited ICa more potently in cells expressing the D40 variant, with morphine also being more potent at D40 in sympathetic neurons (Lopez Soto & Raingo, 2012;

Margas et al., 2007, Table 2). Interestingly, the potency of endomorphin 1 and M-6-G was not different between N40 and D40 in sympathetic neurons. Although the relative expression levels of each receptor were not determined in these studies, the selective enhancement of DAMGO and morphine coupling to ICa in sympathetic neurons suggest that there may be genuine differences in N40 and D40 signalling via this pathway. By contrast, trigeminal neuron ICa from “humanized” N40 and D40 mice was inhibited in an essentially identical manner by DAMGO (Ramchandari et al., 2011) and fentanyl

(Mahmoud et al., 2011), but morphine was less potent and had a lower efficacy in the neurons from the D40 mice (Mahmoud et al., 2011). These results are essentially opposite to those found in the acutely transfected cell lines. There is no ready explanation for these

42 differences, although differences in receptor expression cannot be ruled out. It is likely that HEK293 cells, rat sympathetic neurons and mouse trigeminal neurons express different complements of Gα and βγ subunits, which may also contribute to observed differences. Finally, it should be noted that the humanized N40/D40 MOPr are hybrids, with only the first exon of the human receptor inserted into mouse, meaning that the receptors are human/mouse chimeras. The receptors had a similar affinity for DAMGO

(Ramchandari et al., 2011), but their signalling properties have not been well characterized.

Deb et al. (2010) expressed N40 and D40 variants of the µ-opioid receptor in the mouse neuroblastoma cell line Neuro2A, with radioligand binding experiments indicating similar levels of receptor expression. Using measurements of protein kinase A (PKA) activity and phosphorylated ERK1/2 (pERK) levels in response to a single concentration of morphine

(1 µM) applied for 5 minutes or 6 days, the investigators found differences in PKA and pERK levels between N40 and D40 expressing cells after 6 days only. Unfortunately, the basal PKA activity and acute agonist-stimulated ERK phosphorylation differed significantly between the cell lines, making sensible interpretations difficult. The differences in the signalling responses of the cells could be due to expression of the different opioid receptor variants, or could have arisen due to variations in the phenotype of different Neuro2A cells at the time of clonal selection.

A6V

The A6V variant is located at the N-terminus of MOPr (Table 1). A6V is quite common in some populations but not others, having been reported at allelic frequencies ranging from less than 1% in Caucasian and east Asian populations (Rommelspacher et al., 2001; Tan et al., 2003) to upwards of 20% in African-American and northern Indian populations

43 (Crowley et al., 2003; Kapur et al., 2007). Few studies have investigated the clinical effects of this polymorphism. Crystal et al. (2012) reported an association between the T/T genotype in African-American women and the risk of alcohol, cocaine, tobacco but not opioid use. Other studies have demonstrated a trend towards a higher frequency of V6 in individuals with substance abuse, however these studies have lacked sufficient statistical power due to small sample sizes, and the confounding factor of overall genotype

(Berrettini et al., 1997; Rommelspacher et al., 2001; Compton et al., 2003; Crowley et al.,

2003).

There are no studies comparing acute A6 and V6 signalling on the predominant isoform of

MOPr. In an assay of cAMP-dependent gene transcription, no difference in potency was found for DAMGO, endomorphin-1 or leu-enkephalin in HEK-293 cells expressing A6 and V6 (Table 2, Fortin et al., 2010). The A6V variant was studied on the MOR1A splice variant sequence expressed in HEK 293 cells, where DAMGO but not morphine showed a higher maximum effect at V6- than A6-MOR1A in an assay of intracellular Ca release catalyzed by a transiently transfected chimeric G protein. No differences in internalisation of the V6-MOR1A receptor in response to DAMGO and morphine were observed

(Ravindranathan et al., 2009). The significance of these findings for more naturalistic coupling of MOPr are unclear, but suggest that further work is warranted.

S42C, D51N, G63V and S66F

Other rare polymorphisms within the N-terminal domain of MOPr have been identified within the population but no clinical or phenotypic information is available (Table 1). The

S42C variant resulted in reduced receptor expression and coupling to intracellular calcium release when assayed on the MOR1A splice variant background (Ravindranathan et al,

2009, Table 2)..

44 Several extracellular domain polymorphisms for which there is no phenotypic data were identified on the GPCR Natural Variant database (Kazius et al. 2007) and examined in a cAMP-dependent gene transcription assay (Fortin et al., 2010). Neither D51N nor G63V showed any differences to wild type MOPr in this assay. However, the S66F variant showed a reduction in the potency of DAMGO and endomorphin-1, but not leu-enkephalin to inhibit AC (Table 2; Fortin et al., 2010).

Transmembrane Domain SNPs

L85I (TM1)

The transmembrane helices of MOPr are important elements of the ligand-binding pocket, and they are essential for transmitting information from the extracellular surface to the intracellular signalling domains and also participate in the formation of oligomers

(Serohijos et al., 2011; Manglik et al. 2012). The L85I variant, in TM1 (Table 1), was first reported by Ravindranathan et al (2009). Although there is no information about the phenotype of people carrying the I85 allele, it has an interesting functional profile in vitro.

Both DAMGO and morphine have a moderately lower efficacy in signalling assays measuring I85 (or I83 – the rat ortholog) activity, however morphine displays an enhanced capacity to promote internalisation of the I85/I83 variant (Ravindranathan et al., 2009;

Cooke et al., 2014). Co-expression of the I85 and L85 receptors results in morphine promoting internalisation of both variants, suggesting that they may form functional dimers (Ravindranathan et al., 2009).

Previous studies have shown that WT-MOPr internalizes relatively poorly in response to morphine, and there is also evidence that high efficacy agonists such as DAMGO and appear to induce receptor desensitisation by different mechanisms to partial agonists such as morphine (Ueda et al., 2001; Borgland et al., 2003; Johnson et al., 2006;

45 Kelly et al., 2008; Bailey et al. 2009). Intriguingly, while morphine activated MOPr has been shown to be a poor substrate for G protein-coupled receptor kinase (GRK) subtypes

2/3 phosphorylation, which is required for endocytosis (Doll et al., 2012), internalisation of I83 MOPr was significantly attenuated with overexpression of a GRK2 dominant negative mutant, suggesting this variant is better able to recruit GRK2 (Cooke et al., 2014).

Hierarchical, multi-site phosphorylation is required for efficient MOPr endocytosis (El

Kouhen et al., 2001) and while morphine induces phosphorylation of the S375 residue on the C-terminus of MOPr, DAMGO also efficiently stimulates phosphorylation of T370 after S375 phosphorylation, resulting in receptor internalisation (Doll et al., 2011,

Grecksch et al., 2011). Morphine stimulated internalisation of I83 was not due to enhanced phosphorylation of S375 compared with WT MOPr, but T370 phosphorylation was not investigated (Cooke et al., 2014). The observations that I83/85 MOPr show apparently decreased signalling efficacy compared with enhanced receptor trafficking in response to morphine underscore the likelihood that distinct receptor conformations underlie each of these processes. It will be interesting to see whether it is possible to further define the structural elements in the region of L85 that are involved in MOPr signalling and phosphorylation, and whether it will be possible to independently manipulate these properties of the agonist/receptor complex.

Compensatory changes in cell signalling processes are associated with chronic MOPr activation, one of the most well described of these is upregulation of AC activity that results in “superactivation” of AC upon opioid withdrawal (Sharma et al., 1975; Avidor-

Reiss et al., 1996; Whistler et al., 1999). It has also been suggested that these compensatory changes are limited by agonist-induced receptor internalisation (Wang et al.,

2003). Ravindranathan et al. tested the I85 variant for changes AC superactivation. Cells expressing L/I85 MOPr variant were chronically treated with morphine for 14h. Upon

46 morphine withdrawal, cells expressing I85 MOPr showed a significantly lower level of cAMP compared with WT cells (2.5 fold and 1.5 fold cAMP levels of morphine naive cells respectively). Upon a 4hr “acute” rechallenge with 10 nM morphine, cAMP levels were again significantly lower in the I85 expressing cells, indicating a reduction AC superactivation and morphine tolerance.

S147C and N152D (TM3)

Computational modeling and x-ray crystallography studies have shown transmembrane domains (TM) 3, 5 and 6 to be of particular importance in the formation of the ligand- binding pocket of MOPr (Serohijos et al., 2011; Manglik et al., 2012). Two polymorphisms in TM3 have been detected within the population, S147C and N152D

(Table 1), both occurring at frequencies of < 1% (Bergen et al., 1997; Uhl et al., 1999;

Befort et al., 2000; Ravindranathan et al., 2009). No information on the clinical phenotype of C147 or D152 carriers is available, and limited functional studies have been published.

When expressed on the MOR1A splice variant backbone C147 appeared to support an increased efficacy and potency for DAMGO and morphine to stimulate intracellular calcium release when compared with S147 (Ravindranathan et al. 2009), however, when expressed on the wild type MOPr backbone, C147 was modestly less effective at supporting agonist-mediated inhibition of cAMP-dependent gene transcription (Fortin et al. 2010). Whether this divergence is because of the different receptor backgrounds or whether it hints at a reciprocal change in the capacity of MOPr to activate different signalling pathways remains unknown. The N152D SNP appears to have reduced affinity for morphine but not opioid peptides, unfortunately it was not possible to measure receptor activity, apparently due to low overall expression (Befort et al., 2001).

47 Table 2: Summary of the key findings about MOPr SNP signalling. * P < 0.05, from original publications. Abbreviations: β-End, β-endorphin; CaM, calmodulin; DAM, DAMGO, [D-Ala2, NMe-Phe4, Gly-ol5]-enkephalin; End-1, endomorphin 1; Fent, fentanyl; L-ENK, [Leu]5enkephalin; Meth, methadone; Mor, morphine; MOR-1A, µ-opioid receptor 1A splice variant; M-6-G, morphine-6-glucuronide; PKA, protein kinase A; SCG, superior cervical ganglion

MOPr Variant Key Observations pEC50 WT pEC50 SNP Bmax WT Bmax Var Reference Unchanged agonist affinity. Similar DAMGO 7.0 - DAM 6.7 - DAM 5.5 6.1 Befort et al., 2001 stimulated GTPγS activation. pmol/mg pmol/mg

Similar cAMP inhibition. Reduced D40 9.1 - MOR 9.0 - MOR 4.8 0.63 Beyer et al., 2004 expression. 8.8 - M-6-G 8.8 - M-6-G pmol/mg pmol/mg N40D 7.9 - β-end 7.8 - β-end 3 x increased β-endorphin affinity for D40 Not provided Not provided Not Not Bond et al., 1998 than WT, and 3x increased potency for provided provided GIRK activation in D40 expressing Xenopus oocytes . Different N/D40 stimulated PKA activity and N/A (1 µM N/A (1 µM 835 830 Deb et al. 2010 ERK1/2 phosphylation after chronic morphine only) morphine fmol/mg fmol/mg morphine treatment. only)

Similar inhibition of cAMP-stimulated CRE 8.8 - DAM 8.8 - DAM Values Similar to Fortin et al., 2010 transcription. 8.8 - End-1 8.9 - End-1 not WT 8.4 - L-Enk 8.5 - L-Enk provided Decreased agonist potency to inhibit AC in 8.6 - DAM 8.1 - DAM* Not 66% of Kroslak et al., 2007 D40-HEK-293 and D40-AV-12 cells 8.4 - Mor 7.8 - Mor* provided WT 8.3 - Meth 7.8 - Meth* 8.4 - β-End 8.1 - β-End

Decreased morphine potency for ICa 7.3 - Mor 6.6 - Mor* Not Similar to Mahmoud et al., 2011 inhibition in mouse trigeminal ganglion cells 7.2 - Fent 7.0 - Fent* provided WT expressing “humanized” D40.

Increased DAMGO and morphine potency 7.5 - DAM 7.8 - DAM* Not Not Margas et al., 2007 for ICa inhibition at D40 expressing rat SCG 7.1 - Mor 7.4 - Mor* provided provided cells. 7.1 - M-6-G 7.1 - M-6-G 7.1 - End-1 7.1 - End-1

48 Decreased D40 expression in SII region of 5.9 - DAM 6.0 - DAM 97 114 Oertel et al., 2009 cortex in post-mortem brain.. SII region- fmol/mg fmol/mg specific decrease in DAMGO efficacy in D40 carriers. No difference in DAMGO potency at D40 N/A N/A Not Similar to Ramchandari et al., 2011

for ICa inhibition in "humanized" mouse provided WT N40D (cont.) trigeminal ganglion cells.

Increased DAMGO potency for Cav2.2 8.6 - DAM 9.5 - DAM* Not Not Soto & Raingo, 2012 inhibition in D40 HEK-293 cells. provided provided Lower mRNA levels of G118 allele for N/A N/A Not Not Zhang et al., 2005 heterozygous A118G carriers in post- provided provided mortem brain. Decreased G118 mRNA and 10-fold decreased D40 expression in CHO- K1 cells. Similar inhibition of cAMP-stimulated CRE 8.8 - DAM 8.6 - DAM Not Similar to Fortin et al., 2010 transcription. 8.8 - End-1 8.7 - End-1 provided WT 8.4 - L-Enk 8.2 - L-Enk A6V Unchanged agonist efficacy and potency for 7.5 - DAM 7.9 - DAM 5.6 5.8 Ravindranathan et al., 2009 intracellular Ca release at A/V6 on MOR-1A 7.4 - Mor 7.3 - Mor pmol/mg pmol/mg backbone.

Decreased agonist potency for intracellular 7.5 - DAM > 6.8 - DAM* 2.7 5.8 Ravindranathan et al., 2009 S42C Ca release at C42 on MOR-1A backbone. 7.4 - Mor > 6.8 - Mor* pmol/mg pmol/mg

Similar inhibition of cAMP-stimulated CRE 8.8 - DAM 8.6 - DAM Not Similar to Fortin et al., 2010 D51N transcription. 8.8 - End-1 8.8 - End-1 provided WT 8.4 - L-Enk 8.4 - L-Enk Similar inhibition of cAMP-stimulated CRE 8.8 - DAM 9.0 - DAM Not Similar to Fortin et al., 2010 G63V transcription. 8.8 - End-1 8.9 - End-1 provided WT 8.4 - L-Enk 8.5 - L-Enk Decreased potency of DAMGO and 8.8 - DAM 8.2 - DAM* Not Similar to Fortin et al., 2010 S66F endormorphin-1 at F66 for inhibition of 8.8 - End-1 8.3 - End-1* provided WT cAMP-stimulated CRE transcription. 8.4 - L-Enk 7.7 - L-Enk*

49 Increased morphine stimulated endocytosis 6.7 - DAM 6.5 - DAM 1.8 2.7 Cooke et al., 2014 in I83-HEK293 cells. Decreased agonist 6.7 - Mor 6.9 - Mor pmol/mg pmol/mg L85I (L83I) efficacy in inhibition of AC and ERK phosphorylation. Increased morphine stimulated endocytosis 7.5 - DAM 7.9 - DAM 5.6 5.2 Ravindranathan et al., 2009 in I85-HEK293 cells. Increased AC 7.4 - Mor 7.7 - Mor pmol/mg pmol/mg L85I superactivation in I85 HEK-293 cells. No change in agonist potency. Decreased agonist potency for inhibition of 8.8 - DAM 8.3 - DAM* Values Similar to Fortin et al., 2010 cAMP-stimulated CRE transcription. 8.8 - End-1 8.4 - End-1* not WT S147C 8.4 - L-Enk 7.9 - L-Enk* provided Increased agonist potency for intracellular 7.5 - DAM 7.9 - DAM* 5.6 5.0 Ravindranathan et al., 2009 Ca release at C147 on MOR-1A backbone. 7.4 - Mor 8.3 - Mor* pmol/mg pmol/mg

Decrease in morphine affinity for D152 in N/A N/A 5.5 1.9 Befort et al., 2001 N152D COS cells. pmol/mg pmol/mg HEK-293 cells expressing C181 failed to 7.5 - DAM N/A 5.6 3.5 Ravindranathan et al., 2009 R181C signal via DAMGO or morphine. 7.4 - Mor pmol/mg pmol/mg Decreased K190 expression in HEK-293 Not provided Not provided Not N/A Fortin et al., 2010 cells. Treatment with naloxone and provided N190K naltrexone both increased K190 expression and inhibition of cAMP-stimulated CRE transcription. Decreased agonist potency at F192 for 7.5 - DAM > 6.8 - DAM* 5.6 4.4 Ravindranathan et al., 2009 intracellular calcium release in HEK-293 7.4 - Mor > 6.8 - Mor* pmol/mg pmol/mg N192F cells expressing F192 on MOR-1A backbone. Decreased basal GTPγS activity at H260 in 8.4 - Mor 8.6 - Mor 3.5 3.9 Wang et al., 2001 HEK-293 cells. Slight decrease in morphine pmol/mg pmol/mg R260H stimulated GTPγS at H260, and slight decrease in affinity of H260 for CaM.

50 Decreased basal GTPγS activity at H260 in 7.0 - DAM 6.9 - DAM 5.5 4.6 Befort et al., 2001 COS cells. Slight decrease in maximal pmol/mg pmol/mg DAMGO stimulated GTPγS at H260. Decreased agonist potency for inhibition of 8.8 - DAM 8.0 - DAM* Values Similar to Fortin et al., 2010 cAMP-stimulated CRE transcription. 8.8 - End-1 8.1 - End-1* not WT R265H 8.4 - L-Enk 7.6 - L-Enk* provided Decreased basal GTPγS activity at H265 in 8.4 - Mor 8.5 - Mor 3.5 4.2 Wang et al., 2001 HEK-293 cells. Slight decrease in maximal pmol/mg pmol/mg morphine stimulated GTPγS at H265. Decreased affinity of H265 for CaM binding, and decreased desensitisation following morphine pretreatment. No of agonist-stimulated GTPγS binding in 7.2 - DAM 6.4 - DAM* 5.5 3.6 Befort et al., 2001 COS cells. Decreased agonist potency and 6.5 - β-End 5.9 - β-End* pmol/mg pmol/mg efficacy at P268 for inhibition of cAMP 6.2 - Mor 5.8 - Mor* accumulation. Decreased potency of DAMGO and 8.8 - DAM 8.2 - DAM* Values Similar to Fortin et al., 2010 endomorphin-1 for inhibition of cAMP- 8.8 - End-1 8.4 - End-1* not WT S268P stimulated CRE transcription. 8.4 - L-Enk 7.9 - L-Enk provided

Decreased GTPγS binding, slower N/A N/A 643 340 Koch et al., 2000 desensitisation and decreased AC inhibition fmol/mg fmol/mg in response to DAMGO. Decreased morphine potency at P268 for 7.0 - Mor 6.3 - Mor* 3.5 4.5 Wang et al., 2001 inhibition of cAMP accumulation. pmol/mg pmol/mg

Increased agonist potency for inhibition of 8.8 - DAM 9.1 - DAM* Values Similar to Fortin et al., 2010 D274N cAMP-stimulated CRE transcription. 8.8 - End-1 8.3 - End-1* not WT 8.4 - L-Enk 8.6 - L-Enk* provided Unchanged in agonist potency for inhibition 8.8 - DAM 8.8 - DAM Values Similar to Fortin et al., 2010 V293I of cAMP accumulation. 8.8 - End-1 8.8 - End-1 not WT 8.4 - L-Enk 8.4 - L-Enk provided

51 C192F (TM4)

One SNP in TM4 of OPRM1 has been identified; C192F (Ravindranathan et al. 2009).

When expressed on the MOR1A splice variant backbone, C192F showed significant decreases in DAMGO and morphine potency to mobilize calcium in HEK293 cells transfected with an engineered G protein. No phenotypic information is available.

V293I (TM6)

Shi et al. (2002) detected a V293I amino acid exchange in MOPr. I293 was reported to signal in an identical manner to V293 (Fortin et al., 2010) and there is no clinical information about this phenotype.

Intracellular Loop SNPs

The intracellular loop (ICL) domains of MOPr form major elements of the cytoplasmic interface between the receptor and intracellular effector proteins. ICL2 and 3 have been shown to be of key importance in G-protein coupling of GPCRs, as well as being involved in regulatory processes such as receptor phosphorylation, uncoupling and internalisation

(Lefkowitz, 1998). GPCRs with the ICL2 and ICL3 domains deleted cannot couple to G- proteins but can retaining their ligand-binding properties, and there are a number of examples of ICL SNPs affecting selectivity of G-protein coupling (Capeyrou et al., 1997;

Visiers et al., 2001; Goldfeld et al., 2011; Zheng et al. 2013).

ICL2 contains the highly conserved E/DRY motif, mutations in which have been shown to reduce MOPr agonist efficacy, and also to increase constitutive activity of MOPr (Li et al.,

2001; Clayton et al., 2010). Mutations in ICL2 have also been shown to affect receptor uncoupling and desensitisation (Celver et al., 2001; Celver et al., 2004). ICL3 is highly

52 conserved among all opioid receptor types and has been shown to be involved in basal and agonist-stimulated G-protein coupling, β-arrestin recruitment and contains multiple phosphorylation consensus sequences (Merkouris et al., 1996; Georgoussi et al., 1997;

Wang, 1999b). Mutations within the ICL3 of MOPr have been shown to differentially affect agonist potency and efficacy (Chaipatikul et al., 2003). In addition to their role in acute signalling and short-term regulatory processes, the intracellular domains of GPCRs may be of importance in long-term adaptations to chronic opioid exposure, and contribute to the development of opioid tolerance (Chavkin et al., 2001; Koch & Hollt, 2008;

Williams et al., 2013).

R181C (ICL2)

The R181C variant was reported by Ravindranathan et al. (2009). Interestingly, C181 appears to have an unchanged affinity for DAMGO, but it fails to promote calcium mobilisation or be internalized in response to agonist. Whether the receptor is unable to signal to all effectors remains to be established.

N190K (ICL2)

The rare N190K variant is located at the base of TM4, and was originally reported as an

ICL2 SNP (Table 1; Fortin et al. 2010). Total K190 expression in HEK293 cells is lower than N190, but cell surface expression is almost absent. DAMGO fails to signal through

K190, although it is not clear if this is because of the inaccessibility of the intracellular receptor or a change in the transduction of peptide agonist signals. Interestingly, treatment of the K190 variant with the non-peptide, cell permeable opioid receptor ligands naltrexone, naloxone, buprenorphine or β-CNA (10 μM, 18h) increased cell surface receptor expression, with naltrexone treatment producing levels similar to WT-MOPr.

Small, membrane permeable ligands can “rescue” misfolded or immature GPCR, including

53 opioid receptors, by stabilizing a more native-type conformation in the endoplasmic reticulum and allowing the protein to enter the appropriate secretory pathway (Petaja-Repo et al., 2002; Ulloa-Aguirre et al., 2004; Chen et al., 2006). Fascinatingly, naloxone and naltrexone were also apparently agonists at K190, producing significant inhibition of cAMP-stimulated reporter gene transcription after prolonged treatment (Fortin et al.,

2010). This suggests that K190 is not misfolded/misconfigured to such a degree that it cannot recognize G proteins, but that it nonetheless has an abberant native conformation.

Four rare, naturally occurring SNPs present on ICL3 have been described, R260H (Bond et al., 1998), R265H (Hoehe et al., 2000; Befort et al.; 2001; Wang et al., 2001), S268P

(Hoehe et al., 2000) and D274N (Wang et al., 2001; Table 1). The importance of ICL3 in the regulation and signalling of MOPr has prompted investigation of the functional consequences of ICL3 polymorphisms, despite their rarity within the population.

R260H, R265H, S268P

The R260H and R265H variant receptors exhibited very similar ability to bind opioids and be activated by morphine or DAMGO, with minor differences in agonist-stimulated

GTPγS binding potentially accounted for by small differences in receptor expression or the proportion of receptors on the cell surface. An intriguing finding was that basal GTPγS activity was significantly lower in cells expressing H260 or H265, suggesting a lower constitutive activity (Befort et al, 2001; Wang et al., 2001).

Assays of cAMP accumulation have provided inconsistent results with respect to H260 or

H265 signalling. Wang et al. (2001) found no differences in morphine potency or efficacy for inhibition of forskolin-stimulated radiolabelled cAMP accumulation in cells expressing

WT-MOPr, H260 or H265 while Befort et al. (2001) also found no differences between

54 H265 and WT in a cAMP response element (CRE) reporter gene assay (see Table 2). By contrast, Fortin et al. (2010) used a different CRE reporter assay and found a decrease in potency of DAMGO, endomorphin-1 and leu-enkephalin signalling through both H260 and

H265. It is difficult to directly compare these studies as Fortin et al. (2010) did not quantify receptor expression, but the relatively high levels of receptor expression in the cells used by Wang et al. (2001) and Befort et al. (2001) could conceivably reduce the sensitivity of the assay to detect differences in agonist potency at the variant receptors.

A third ICL3 variant, S268P, results in the loss of a putative Ca2+/calmodulin-dependent protein kinase II (CaMK II) phosphorylation site (Koch et al., 2000) and insertion of an amino acid, proline, that is likely to significantly disrupt the structure of ICL3. Most studies have found that P268 or the rat ortholog S266P (Koch et al., 2000) have a significantly reduced signalling capacity, although the extent of this depends somewhat on the assay conditions used (Koch et al., 2000, Befort et al. 2001; Wang et al., 2001; Fortin et al. 2010; Table 2). The reduction in signalling does not seem to be associated with a change in ligand affinity for the receptor (Koch et al., 2000; Befort et al. 2001), but it is unclear what the relative contributions of the loss of the potential phosphorylation site or the introduction of the proline residue are to the observed in vitro phenotype.

Mutations in ICL3 of MOPr affect signalling of the receptor, but changes in the signalling profile of MOPr resulting from ICL3 SNPs are likely to be expressed in situations other than acute MOPr signalling because ICL domains of GPCRs interact with effectors involved in receptor regulation and adaptive processes such as receptor downregulation

(Lefkowitz, 1998). The ICL3 domain of MOPr has multiple consensus phosphorylation sites, as well as a putative calmodulin binding domain (Wang et al., 1999a; Koch et al.

2000). Sustained exposure to high concentrations of agonist produces downregulation of

55 receptor protein in cell lines, and the ICL3 variants R260H, R265H and S268P were downregulated (~ 80%) to a similar degree to WT receptors by 10 µM DAMGO (Befort et al., 2001). Functionally, P268 MOPr-mediated inhibition of AC desensitized with a similar time course to T268, while desensitisation of P268-mediated activation of GIRK was slower and incomplete when compared to T268 when the proteins were expressed in

Xenopus oocytes (Koch et al., 2000).

In addition to phosphorylation sites, MOPr ICL3 contains a putative calmodulin (CaM) binding site. It has been suggested that CaM competes with G-protein coupling at ICL3, and regulates basal MOPr signalling (Wang et al., 1999). Wang et al. (2001) investigated interaction of CaM with MOPr ICL3 domains by incubating CaM with short peptides derived from ICL3 sequences as well as full length MOPr purified from transfected HEK-

293 cells. The ICL3 H260 peptide showed marginal reduction of CaM binding but the

ICL3 H265 and P268 peptides bound CaM significantly less well. A similar pattern was observed in western blots of full-length MOPr variants bound to CaM. The broader significance of these findings have not been firmly established.

D274N

The D274N variant has received much less attention than ICL3 variants discussed earlier.

It was originally reported by Wang et al. (2001), but not investigated until the study of

Fortin et al. (2010). DAMGO and leu-enkephalin showed a slight increase in potency for inhibition of cAMP accumulation at N274, while endomorphin-1 potency was significantly increased when compared with WT-MOPr in HEK-293 cells (see Table 2). No change in

DAMGO efficacy was observed. These results are in direct contrast to other ICL3 variants, where receptor signalling tended to be reduced.

56 Limitations of extant functional SNP studies

The interpretation of studies of opioid receptor function in vitro, and the extent to which fruitful comparisons can be made between studies are subject to several important caveats.

These extend beyond the everyday differences in the way that laboratories perform studies, and can limit the confidence we have in our understanding of the impact non-synonymous

SNPs have on OPRM1 function. Firstly, many studies do not quantify receptor expression, either in the whole cell or on the cell surface. While it is unrealisitic to expect

“physiological” expression levels (whatever they may be) in all expression systems, high levels of receptor can lead to significant receptor reserve or exaggerated coupling to effectors not normally accessed by the receptor. Receptor reserve is an important issue that has apparently rarely been considered, and even modest differences in receptor expression could significantly affect the signalling profile of important partial agonists such as morphine, and spare receptors may mask subtle differences between variant signalling.

Secondly, the techniques used to measure MOPr activation in many studies do not reflect acute, real-time, naturalistic signalling of MOPr. MOPr undergoes rapid desensitisation and internalisation following agonist exposures of 5-10 mins (Connor et al., 2004). Thus, the reporter gene assays used for facile quantification of MOPr function measure the summed effects of MOPr activation, desensitisation, internalisation and resensitisation, and this may obscure differences between variants at any of these points. Clonal selection of transformed cells during establishment of cell lines expressing variants may contribute to signalling differences observed between variants, and this is rarely controlled for with experiments on endogenous GPCR in each cell line used. These shortcomings are common to many studies of cell signalling in heterologous systems, and to an extent come with the

57 territory, but they are especially important to consider and try and minimize given the potentially subtle nature of changes produced by SNPs.

MOPr is expressed in a wide variety of human cell types, and subtle changes in MOPr signalling arising from SNPs are likely to differ between tissue and cell type. As such, it is difficult to lay out an “ideal” strategy for investigating functional consequences of MOPr

SNPs. In reality, studies undertaken in a variety of heterologous expression systems are probably useful for capturing the range of signalling and regulatory differences that may be produced by MOPr variants (e.g. Charfi et al., 2013). However, simple measures that might enable more confident assertions that differences seen might represent more than just an experimental quirk would include using similar expression systems when attempting to make direct comparisons between the effects of changes in MOPr sequence and/or the effects of multiple ligands, controlling for receptor expression and reserve, and examining as many effectors as possible in similar conditions. Practical steps towards this include the use of cell lines with defined recombination sites to allow the construction of multiple clones on an isogenic background (e.g. FlpIn cells, Knapman et al., 2014) and the use of inducible expression systems or transient transfections to minimize the effects of prolonged expression of MOPr on cell phenotype and perhaps gain some ability to titrate the amount of cell surface receptor (e.g. Fortin et al., 2010; Knapman et al., 2014). It is always useful to use assays that capture the kinetics of drug/receptor/second messenger activity, rather than simply endpoint assays (e.g. Johnson et al., 2006; Cawston et al.,

2013; Knapman et al., 2013, 2014; Tudashki et al., 2014) and it is also important to have a system where changes in efficacy can be readily determined, whether by use of pharmacological tools or by choosing cell lines where there are a minimum of spare receptors. Defining receptor reserve using irreversible antagonists such as β- funaltrexamine or β-chlornaltrexamine and then fitting data to operational models (e.g.

58 Borgland et al., 2003; Rivero et al., 2012; Kelly, 2013) can allow for precise determination of rank orders of agonist efficacy and uncover differences in signalling across different effectors in the same cell, enabling a more complete characterisation of the consequences of changes in receptor sequence. All these ideas have been extensively reviewed in the context of defining ligand bias and allostery at GPCRs, and there is no reason they should not be applied when it is the receptor that changes rather than the ligand (Kenakin &

Christopoulos, 2013).

Future Studies

Areas of great importance for opioid receptor function remain largely unexplored for most

SNPs. In particular, the efficiency of coupling of SNPs to the range of possible MOPr signalling pathways has barely been touched on, as have possible ligand-specific changes in this coupling. Several studies have examined the trafficking of MOPr variants in response to a limited range of agonists (Ravindranathan et al., 2009; Cooke et al., 2014), but the effect of MOPr SNPs on the rapid desensitisation of signalling that precedes receptor internalisation remains unknown. The way in which MOPr SNPs may affect the occurrence or function of putative MOPr dimers has received limited attention

(Ravindranathan et al., 2009), even though most carriers of variant MOPr alleles will be heterozygous for the wild type receptor.

Understanding how MOPr SNPs affect cellular signalling is important for predicting the potential clinical or phenotypic consequences of these variants in humans. However, understanding other aspects of MOPr function such as the regulation of gene expression in response to environmental or epigenetic factors, and the function of MOPr in the wide range of human cells which normally express it, are equally important and more difficult to achieve. Nevertheless, understanding the consequences of expressing a particular MOPr

59 variant should one day contribute to a more personalized approach to opioid prescription.

The ability to predict the effects of specific opioid drugs in individuals, including side effects and the development of tolerance, would minimize the risk of serious adverse events associated with , while maximizing therapeutic benefits and ensuring individuals receive adequate pain relief. Such prediction would necessarily involve determining the genotype of multiple proteins involved in opioid ligand distribution and metabolism, as well as effectors downstream of MOPr, but a key element would be knowing what version of MOPr a patient had, and knowing which of the many opioid analgesics available had the best pharmacodynamic profile at that variant.

Acknowledgements

AK supported by a MQRES Scholarship. MC supported by NH&MRC Project Grant

1011979.

References

Abrol R, Kim SK, Bray JK, Trzaskowski B, Goddard WA 3rd (2013). Conformational ensemble view of G protein-coupled receptors and the effect of mutations and ligand binding. Methods Enzymol 520: 31-48.

Alexander SPH, Mathie A, Peters JA (2011) Guide to receptors and channels (GRAC), 5th edn. Brit J Pharmacol 164 (Suppl.1) S1-S324.

60 Atwood BK, Lopez J, Wager-Miller J, Mackie K, Straiker A (2012). Expression of G protein-coupled receptors and related proteins in HEK293, AtT20, BV2, and N18 cell lines as revealed by microarray analysis. BMC Genomics 12:14

Avidor-Reiss T, Nevo I, Levy R, Pfeuffer T, Vogel Z (1996). Chronic opioid treatment induces adenylyl cyclase V superactivation. Involvment of Gβγ. J Biol Chem 271: 21309-

21315.

Bailey CP, Oldfield S, Llorente J, Caunt CJ, Teschemacher AG, Roberts L and others

(2009). Involvement of PKC alpha and G-protein-coupled receptor kinase 2 in agonist- selective desensitisation of mu-opioid receptors in mature brain neurons. Br J Pharmacol

158: 157-164.

Berfort K, Filliol D, Decaillot FM, Gaveriaux-Ruff C, Hoehe MR, Kieffer BL (2001). A single nucleotide polymorphic mutation in the human mu-opioid receptor severely impairs receptor signalling. J Biol Chem 276: 3130-7.

Bergen AW, Kokoszka J, Peterson R, Long JC, Virkkunen M, Goldman D (1997). Mu- opioid receptor gene variants: lack of association with alcohol dependence. Mol Psych 2:

490-494.

Berrettini WH, Hoehe MR, Ferraro TN, DeMaria PA, Gottheil E (1997). Human mu- opioid receptor gene polymorphisms and vulnerability to substance abuse. Addiction Biol

2: 303-308.

61 Beyer A, Koch T, Schroder H, Schulz S, Hollt V (2004). Effect of the A118G polymorphism on binding affinity, potency and agonist-mediated endocytosis, desensitisation and resensitisation of the human mu-opioid receptor. J Neurochem 89: 553-

560.

Bond C, LaForge KS, Tian M, Melia D, Zhang S, Borg L et al. (1998). Single-nucleotide polymorphism in the human mu opioid receptor gene alters β-endorphin binding and activity: Possible implications for opiate addiction. Proc Natl Acad Sci USA 95: 9608-

9613.

Borgland SL, Connor M, Osborne PB, Furness JB, Christie MJ (2003). Opioid agonists have different efficacy profiles for G protein activation, rapid desensitisation, and endocytosis of mu-opioid receptors. J Biol Chem 278: 18776-18784.

Capeyrou R, Riond J, Corbani M, Lepage JF, Bertin B, Emorine LJ (1997). Agonist- induced signalling and trafficking of the μ-opioid receptor: role of serine and threonine residues in the third cytoplasmic loop and C-terminal domain. FEBS Letters 415: 200-205.

Cawston EE, Redmond WJ, Breen C, Grimsey N, Connor M, Glass M (2013). Real-time characterisation of 1 (CB1) allosteric modulators reveals novel mechanism of action. Brit J Pharmacol 170: 893-907.

Celver JP, Lowe J, Kovoor A, Gurevich VV, Chavkin C (2001). Threonin 180 is required for G-protein-coupled receptor kinase 3- and beta-arrestin 2-mediated desensitisation of the mu-opioid receptor in Xenopus oocytes. J Biol Chem 276: 4894-4900.

62 Celver J, Xu M, Jin W, Lowe J, Chavkin C (2004). Distinct domains of the mu-opioid receptor control uncoupling and internalisation. Mol Pharmacol 65: 528-537.

Chaipatikul V, Loh, HH, Law PY (2003). Ligand-selective activation of μ-opioid receptor:

Demonstrated with deletion and single amino acid mutations of third intracellular loop domain. J Pharmacol Exp Ther 305: 909-918.

Chavkin C, McLaughlin JP, Celver JP (2001). Regulation of opioid receptor function by chronic agonist exposure: constitutive activity and desensitisation. Mol Pharmacol 60: 20-

25.

Charfi I, Nagi K, Mnie-Filali O, Thibault D, Balboni G, Schiller PW et al, (2013) Ligand- and cell-dependent determinants of internalisation and cAMP modulation by delta opioid receptor (DOR) agonists. Cell Mol Life Sci, In Press.

Chen Y, Chen C, Wang Y, Liu-Chen LY (2006). Ligands regulate cell surface level of the human kappa opioid receptor by activation-induced down-regulation and pharmacological chaperone-mediated enhancement: differential effects of nonpeptide and peptide agonists.

J Pharmacol Exp Ther 319: 765-775.

Clayton CC, Bruchas MR, Lee ML, Chavkin C (2010). Phosphorylation of the µ-opioid receptor at tyrosine 166 (Tyr3.51) in the DRY motif reduces agonist efficacy. Mol

Pharmacol 77: 339-347.

63 Compton P, Geschwind DH, Alarcon M (2003). Association between human mu-opioid receptor gene polymorphism, pain tolerance, and opioid addiction. Am J Med Genet B

Neuropsychiatr Genet 121B(1): 76-82.

Connor M, Christie MD (1999). Opioid receptor signalling mechanisms. Clin Exp

Pharmacol Physiol 26: 493-499

Connor M, Osborne PB, Christie MJ (2004). Mu-opioid receptor desensitisation: is morphine different? Br J Pharmacol 143: 685-696.

Corbett AD, Henderson G, McKnight AT, Paterson SJ (2006). 75 years of opioid research: the exciting but vain quest for the Holy Grail. Br J Pharmacol 147 Suppl 1: S153-162.

Cox B (2013). Recent developments in the study of opioid receptors. Mol Pharmacol

83:723-728.

Crowley JJ, Oslin DW, Patkar AA, Gottheil E, DeMaria PA Jr, O’Brien CP et al. (2003).

A genetic association study of the mu opioid receptor and severe opioid dependence.

Psychiatr Genet 13: 169-73.

Crystal HA, Hamon S, Randesi M, Cook J, Anastos K, Lazar J (2012). A C17T polymorphism in the mu opiate receptor is associated with quantitative measures of drug use in African American women. Addict Biol 17: 181-91.

Dahan A, Aarts L & Smith TW (2010). Incidence, reversal and prevention of opioid- induced respiratory depression. Anesthesiology 112: 226-238.

64 Dang VC, Christie MJ (2012). Mechanisms of rapid opioid receptor desensitisation, resensitisation and tolerance in brain neurons. Br J Pharmacol 165: 1704-1716.

Deb I, Chakraborty J, Gangopadhyay PK, Choudhury SR, Das S (2010). Single-nucleotide polymorphism (A118G) in exon 1 of OPRM1 gene causes alteration in downstream signalling by mu-opioid receptor and may contribute to the genetic risk for addiction. J

Neurochem 112: 486-96.

Dhalla IA, Persaud N, Juurlink DN (2011). Facing up to the prescription opioid crisis. BMJ

343: d5142.

Diatchenko L, Robinson JE, Maixner W (2011). Elucidation of mu-opioid gene structure:

How genetics can help predict therapeutic response to opioids. Eur J Pain Suppl 5: 433-48.

Doll C, Konietzko J, Poll F, Koch T, Hollt V, Schulz S (2011). Agonist-selective patterns of mu-opioid receptor phosphorylation revealed by phosphosite-specific antibodies. Br J

Pharmacol 164: 298-307.

Doll C, Poll F, Peuker K, Loktev A, Gluck L, Schulz S (2012). Deciphering µ-opioid receptor phosphorylation and dephosphorylation in HEK293 cells. Br J Pharmacol 167:

1259-70.

El Kouhen R, Burd AL, Erickson-Herbrandson, LJ, Chang CY, Law PY, Loh HH (2001).

Phosphorylation of Ser363, Thr370 and Ser375 residues within the carboxyl tail differentially regulates μ-opioid receptor internalisation. J Biol Chem 276: 12774-12780.

65 Fortin JP, Ci L, Schroeder J, Goldstein C, Montefusco MC, Peter I et al. (2010). The μ- opioid receptor variant N190K is unresponsive to peptide agonists yet can be rescued by small-molecule drugs. Mol Pharmacol 78: 837-845.

Georgoussi Z, Merkouris M, Mullaney I, Megaritis G, Carr C, Zioudrou C, Milligan G

(1997). Selective interactions of mu-opioid receptors with pertussis toxin-sensitive G proteins: involvement of the third intracellular loop and the c-terminal tail in coupling

Goldfeld DA, Zhu K, Beuming T, Friesner RA (2011). Successful prediction of the intra- and extracellular loops of four G-protein coupled receptors. Proc Natl Acad Sci USA 109:

9665-9671.

Grecksch G, Just S, Pierstorff C, Imhof AK, Gluck, L, Doll C et al. (2011). Analgesic tolerance to high-efficacy agonists but not to morphine is diminished in phosphorylation- deficient S375A mu-opioid receptor knock-in mice. J Neurosci 31:13890-13896.

Hoehe MR, Kopke K, Wendel B, Rohde K, Flachmeier C, Kidd KK and others (2000).

Sequence variability and candidate gene analysis in complex disease: association of mu opioid receptor gene variation with substance dependence (2000). Hum Mol Genet 9:

2895-2908.

Huang P, Chen C, Mague SD, Blendy JA, Liu-Chen LY (2012). A common single nucleotide polymorphism A118G of the μ opioid receptor alters its N-glycosylation and protein stability. Biochem J 441: 379-386.

66 Ikeda K, Ide S, Han W, Hayashida M, Uhl GR, Sora I (2005). How individual sensitivity to can be predicted by gene analyses. Trend Pharmacol Sci 26: 311-317.

Johnson EA, Oldfield S, Braksator E, Gonzalez-Cuello A, Couch D, Hall KJ (2006).

Agonist-selective mechanisms of mu-opioid receptor desensitisation in human embryonic 293 cells. Mol Pharmacol 70: 676-685.

Jordan B, Devi LA (1998). Molecular mechanisms of opioid receptor signal transduction.

Brit J Anaesth 81: 12-19.

Kapur S, Sharad S, Singh RA, Gupta AK (2007). A118G polymorphism in mu opioid receptor gene (oprm1): association with opiate addiction in subjects of Indian origin. J

Integr Neurosci 6: 511-522.

Kelly E, Bailey CP, Henderson G (2008). Agonist-selective mechanisms of GPCR desensitisation. Br J Pharmacol 153 Suppl 1: S379-388.

Kelly (2013). Efficacy and ligand bias at the µ-opioid receptor. Brit J Pharmacol 169:

1430-1446.

Kenakin T (2002). Efficacy at G-protein-coupled receptors. Nat Rev Drug Discov 1: 103-

110

Kenakin T, Miller LJ (2010). Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery.

Pharmacol Rev 62: 265-304.

67 Kenakin T, Christopoulos A (2013). Signalling bias in new drug discovery: detection, quantification and therapeutic impact. Nat Rev Drug Disc, 12: 205-216.

Knapman A, Santiago M, Du Y-P, Bennallack P, Christie MJ, Connor M (2013). A continuous, fluorescence based assay of μ-opioid receptor activation in AtT-20 cells. J

Biomol Screen, 18: 269-276.

Knapman A, Abogadie F, McIntrye P, Connor M (2014). A real time, fluorescence-based assay for measuring µ-opioid receptor modulation of adenylyl cyclase activity in Chinese

Hamster Ovary cells. J of Biomol Screen, 19: 223-231.

Koch T, Hollt V (2008). Role of receptor internalisation in opioid tolerance and dependence. Pharmacol Ther 117: 199-206.

Koch T, Kroslak T, Averbeck M, Mayer P, Schroder H, Raulf E, Hollt V (2000). Allelic variation S268P of the human mu-opioid receptor affects both desensitisation and G protein coupling. Mol Pharmacol 58: 328-334.

Kreek MJ, Bart G, Lilly C, LaForge KS, Nielsen DA (2005). Pharmacogenetics and

Human Molecular Genetics of Opiate and Cocaine and Their Treatments.

Pharmacol Rev 57: 1-26.

Kroslak T, LaForge KS, Gianotti RJ, Ho A, Nielsen DA, Kreek MJ (2007). The single nucleotide polymorphism A118G alters functional properties of the human mu opioid receptor. J Neurochem 103: 77-87.

68 LaForge KS, Yuferov V, Kreek MJ (2000). Opioid receptor and peptide gene polymorphisms: potential implications for addictions. Eur J Pharmacol 410: 249-268.

Law PY, Wong YH, Loh HH (2000). Molecular mechanisms and regulation of opioid receptor signalling. Annu Rev Pharmacol Toxicol 40: 389-430.

Lefkowitz RJ (1998). G protein couple receptors III. New roles for receptor kinases and β- arrestins in receptor signalling and desensitisation. J Biol Chem 273: 18677-18680.

Li J, Huang P, Chen C, de Riel JK, Weinstein H, Liu-Chen LY (2001). Constitutive activation of the mu-opioid receptor by mutation of D3.49(164), but not D3.32(147):

D3.49(164) is critical for stabilisation of the inactive form of the receptor and for its expression. Biochemistry 40: 12039-12050.

Lopez Soto EJ, Raingo J (2012). A118G mu opioid receptor polymorphism increases inhibitory effects on CaV2.2 channels. Neurosci Lett 523: 190-194.

Lotsch J, Geisslinger G (2005). Are μ-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol Med 11: 82-89.

Mague SD, Isiegas C, Huang P, Liu-Chen LY, Lerman C, Blendy JA (2009). Mouse model of OPRM1 (A118G) polymorphism has sex-specific effects on drug-mediated behavior.

Proc Natl Acad Sci USA 106: 10847-10852.

Mague SD, Blendy JA (2010). OPRM1 SNP (A118G): involvement in disease development, treatment response, and animal models. Drug Alcohol Depend 108: 172-182.

69 Mahmoud S, Thorsell A, Sommer WH, Heilig M, Holgate JK, Bartlett SE, Ruiz-Velasco V

(2011). Pharmacological consequence of the A118G μ opioid receptor polymorphism on morphine- and fentanyl-mediated modulation of Ca2+ channels in humanized mouse sensory neurons. Anesthesiology 115: 1054-1062.

Manglik A, Kruse AC, Kobilka TS, Thian FS, Mathiesen JM, Sunahara RK et al. (2012).

Crystal structure of the μ-opioid receptor bound to a morphinan antagonist. Nature 485:

321-326.

Margas W, Zubkoff I, Schuler HG, Janicki PK, Ruiz-Velasco V (2007). Modulation of

Ca2+ channels by heterologously expressed wild-type and mutant human μ opioid receptors (hMORs) containing the the A118G single-nucleotide polymorphism. J

Neurophysiol 97: 1058-1067.

Massotte D, Brillet K, Kieffer BL, Milligan G (2002). Agonists activate Gi1α or Gi2α fused to the human mu opioid receptor differently. J Neurochem 81: 1372-82.

Matthes HWD, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I et al. (1996).

Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the µ-opioid-receptor gene. Nature 383: 819-823.

Mazarakou G, Georgoussi Z (2005). STAT5A interacts with and is phosphorylated upon activation of the mu-opioid receptor. J Neurochem 93: 918-931.

Merikangas KR, Stolar M, Stevens DE, Goulet J, Preisig MA, Fenton B et al. (1998).

Familial transmission of substance use disorders. Arch Gen Psychiatry 55: 973-9.

70 Merkouris M, Dragatsis I, Megaritis G, Konidakis G, Zioudrou C, Milligan G, Georgoussi

Z (1996). Identification of the critical domains of the delta-opioid receptor involved in G protein coupling using site-specific synthetic peptides. Mol Pharmacol 50: 985-993.

Mizoguchi H, Watanabe C, Sakurada T, Sakurada S (2012). New vistas in opioid control of pain. Curr Opin Pharmacol 12(1): 87-91.

Moore RA, McQuay HJ (2005). Prevalence of opioid adverse events in chronic non- malignant pain: systematic review of randomized trials of oral opioids. Arthrit Res Ther 7:

R1046-R1051

Mura E, Govoni S, Racchi M, Carossa V, Ranzani GN, Allegri M, van Schaik RHN

(2013). Consequences of the 118A>G polymorphism in the OPRM1 gene: translation from bench to bedside? J Pain Res 6: 331-53.

Noble M, Treadwell JR, Tregear SJ, Coates VH, Wiffen PJ, Akafomo C et al. (2010).

Long-term opioid management for chronic noncancer pain. Cochrane Database of

Systemic Reviews 1: CD006605

Oertel BG, Kettner M, Scholich K, Renne C, Roskam B, Geisslinger G et al. (2009). A common human μ-opioid receptor genetic variant diminishes the receptor signalling efficacy in brain regions processing the sensory information of pain. J Biol Chem 284:

6530-6535.

71 Oertel BG, Doehring A, Roskam B, Kettner M, Hackmann N, Ferreiros N et al. (2012).

Genetic-epigenetic interaction modulates μ-opioid receptor regulation. Hum Mol Genet 21:

4751-4760.

Pasternak GW, Pan YX (2013) Mu opioids and their receptors: evolution of a concept.

Pharmacol Rev 65: 1257-317.

Petäjä-Repo UE, Hogue M, Bhalla S, Laperrière A, Morello JP, Bouvier M (2002).

Ligands act as pharmacological chaperones and increase the efficiency of delta opioid receptor maturation. EMBO J 21:1628-37.

Pineyro G, Archer-Lahlou E (2007). Ligand-specific receptor states: implications for opiate receptor signalling and regulation. Cell Signal 19: 8-19.

Ramchandani VA, Umhau J, Pavon FJ, Ruiz-Velasco V, Margas W, Sun H et al. (2011). A genetic determinant of the striatal dopamine response to alcohol in men. Mol Psychiatry

16: 809-817.

Ravindranathan A, Joslyn G, Robertson M, Schuckit MA, Whistler JL, White RL (2009).

Functional characterisation of human variants of the mu-opioid receptor gene. Proc Natl

Acad Sci USA 106: 10811-10816.

Rivero G, Llorente J, McPherson J, Cooke A, Mundell S, McArdle CA et al, (2012)

Endomorphin 2: A biased agonist at the µ-opioid receptor. Mol Pharmacol 82: 178-188.

72 Rommelspacher H, Smolka M, Schmidt LG, Samochowiec J, Hoehe MR (2001). Genetic analysis of the mu-opioid receptor in alcohol-dependent individuals. Alcohol 24: 129-135.

Saidak Z, Blake-Palmer K, Hay DL, Northup JK, Glass M (2006). Differential activation of G-proteins by μ-opioid receptor agonists. Brit J Pharmacol 147: 671-680.

Serohijos AWR, Yin S, Ding F, Gauthier J, Gibson DG, Maixner W et al. (2011).

Structural bias for mu-opioid receptor binding and activation. Structure 19: 1683-1690.

Sharma SK, Nirenberg M, Klee WA (1975). Morphine receptors as regulators of adenylate cyclase activity. Proc Natl Acad Sci USA 72: 590-594.

Shi J, Hui L, Xu Y, Wang F, Huang W, Hu G (2002). Sequence variations in the mu- opioid receptor gene (OPRM1) associated with human addiction to heroin. Hum Mutat 19:

459-460.

Singh VK, Bajpai K, Biswas S, Haq W, Khan MY, Mathur KB (1997). Molecular biology of opioid receptors. Neuroimmunomodulation 4: 285-297.

Skorpen F, Laugsend EA (2008). Variable response to opioid treatment: any genetic predictors in sight? Palliat Med 22: 310-327.

Tan EC, Tan CH, Karupathivan U, Yap EP (2003). Mu opioid receptor gene polymorphisms and heroin dependence in Asian populations. Neuroreport 14: 569-572.

73 Thompson MD, Percy ME, McIntyre WB, Cole DE (2008). G protein-coupled receptors disrupted in human genetic disease. Methods Mol Biol 448: 109-137.

Tudashki HB, Robertson DN, Schiller PW, Pineyro G (2014). Endocytic profiles of δ- opioid receptor ligands determine the duration of rapid but not sustained cAMP responses.

Mol Pharmacol 85: 148-161.

Ueda H, Inoue M, Matsumoto T (2001). Protein kinase C-mediated inhibition of mu-opioid receptor internalisation and its involvement in the development of acute tolerance to peripheral mu-agonist analgesia. J Neurosci 21: 2967-2973.

Uhl GR, Sora I, Wang Z (1999). The mu opiate receptor as a candidate gene for pain: polymorphisms, variations in expression, nociception, and opiate responses. Proc Natl

Acad Sci USA 96: 7752-7755.

Ulloa-Aguirre A, Janovick JA, Brothers SP, Conn PM (2004). Pharmacologic rescue of conformationally-defective proteins: implications for the treatment of human disease.

Traffic 5: 821-37.

Visiers I, Hassan SA, Weinstein H (2001). Differences in conformational properties of the second intracellular loop (IL2) in 5HT(2C) receptors modified by RNA editing can account for G protein coupling efficiency. Protein Eng 14: 409-414.

Walter C, Lotsch J (2009). Meta-analysis of the relevance of the OPRM1 118A>G genetic variant for pain treatment. Pain 146: 270-275.

74 Wang D, Quillan JM, Winans K, Lucas JL, Sadee W (2001). Single nucleotide polymorphisms in the human µ opioid receptor gene alter basal G protein coupling and calmodulin binding. J Biol Chem 276: 34624-34630.

Wang D, Sadee W, Quillan JM (1999). Calmodulin binding to G protein coupling domain of opioid receptors. J Biol Chem 274: 22081-22088.

Wang HL (1999). A conserved arginine in the distal third intracellular loop of the mu- opioid receptor is required for G protein activation. J Neurochem 72: 1307-14.

Wang Y, Li JG, Huang P, Xu W, Liu-Chen LY (2003). Differential effects of agonists on adenylyl cyclase superactivation mediated by the kappa opioid receptors: adenylyl cylcase superactivation is independent of agonist-induced phosphorylation, desensitisation, internalisation, and down-regulation. J Pharmacol Exp Ther 307: 1127-1134.

Wang YJ, Huang P, Ung A, Blendy JA, Liu-Chen LY (2012). Reduced expression of the

µ-opioid receptor in some, but not all, brain regions in mice with OPRM1 A112G.

Neuroscience 205: 178-184.

Whistler JL, Chuang HH, Chu P, Jan LY, von Zastrow M (1999). Functional dissociation of mu opioid receptor signalling and endocytosis: implications for the biology of opiate tolerance and addiction. Neuron 23: 737-746.

Williams JT, Ingram SL, Henderson G, Chavkin C, von Zastrow M, Schulz S, et al (2013).

Regulation of µ-opioid receptors: desensitisation, phosphorylation, internalisation, and tolerance. Pharmacol Rev 65: 223-254.

75

Zhang F, Steinberg SF (2013). S49G and R389G polymorphisms of the β- influence signalling via the cAMP-PKA and ERK pathways. Physiol Genomics

(Epub ahead of print).

Zhang L, Loh HH, Law PY (2013). A novel noncanonical signalling pathway for the µ- opioid receptor. Mol Pharmacol 84: 844-853.

Zhang Y, Wang D, Johnson AD, Papp AC, Sadee W (2005). Allelic expression imbalance of human mu opioid receptor (OPRM1) caused by variant A118G. J Biol Chem 280:

32618-24.

Zheng C, Chen L, Chen X, He X, Yang J, Shi Y, Zhou N (2013). The second intracellular loop of the human cannabinoid CB2 receptor governs G protein coupling in coordination with the carboxyl terminal domain. PLoS One 8: e63262.

Zheng H, Loh HH, Law PY (2010). Agonist-selective signalling of G protein-coupled receptor: mechanisms and implications. IUBMB Life 62: 112-119.

76 Hypotheses and Aims

The overall hypothesis of this thesis is that naturally occurring OPRM1 SNPs contribute to the variability observed between individuals in response to opioid analgesics by causing changes in MOPr signalling and function. The distinct aims of this study are to:

1. Determine whether MOPr variants will couple to distinct signalling pathways with

different efficiencies compared with wild-type MOPr.

2. Determine whether opioid ligands will vary in their ability to stimulate distinct

signalling pathways.

3. Determine whether opioid ligands are differentially affected in their ability to stimulate

distinct signalling pathways at MOPr variants compared with wild-type MOPr.

77 METHODS

The next 2 chapters describe two novel high-throughput screening (HTS) assays that were developed as part of this thesis, “A real-time, fluorescence-based assay for measuring μ- opioid receptor modulation of adenylyl cyclase activity in Chinese hamster ovary cells”

(Chapter 3) and “A continuous, fluorescence-based assay of μ-opioid receptor activation in

AtT-20 cells” (Chapter 4). A detailed protocol for these assays is outlined in the Appendix,

“Fluorescence-Based, High-Throughput Assays for µ-Opioid Receptor Activation using

Membrane Potential-Sensitive Dye”. These assays were developed in order to address the aims listed above, by enabling simple and rapid measurement of MOPr signalling via 11 opioid ligands at 6 MOPr variants. In subsequent chapters I explore the consequences of

MOPr SNPs for receptor signalling via a wide range of opioid agonists utilising these assays, along with a whole-cell ELISA assay for ERK1/2 phosphorylation described in

Chapter 5, “Buprenorphine signalling is compromised at the N40D polymorphism of the human µ-opioid receptor in vitro".

78 Chapter 3

A real-time, fluorescence-based assay for measuring μ-opioid receptor modulation of adenylyl cyclase activity in Chinese hamster ovary cells

Alisa Knapman, Fe Abogadie, Peter McIntrye, Mark Connor

This paper was published in 2014 in the Journal of Biomolecular Screening 19: 223-231.

Fe Abogadie and Peter McIntyre constructed the T-Rex FlpIn CHO cells that were subsequently used to express MOPr variants. Mark Connor assisted in experimental design and manuscript preparation. All other work is my own.

79 ABSTRACT

Inhibition of adenylyl cyclase (AC) activity is frequently used to measure µ-opioid receptor (MOR) activation. We sought to develop a simple, rapid assay of AC activity in whole cells that could be used to study MOR signalling. Chinese hamster ovary cells expressing human MOR (CHO-MOR cells) were grown in 96-well plates and loaded with membrane potential-sensitive fluorescent dye. CHO-MOR cells were treated with the AC activator forskolin (FSK), with or without simultaneous application of MOR agonists, and the resulting change in fluorescence was measured. CHO-MOR cells hyperpolarised in response to application of FSK (pEC50 7.3) or calcitonin (pEC50 9.4). A submaximally effective concentration of FSK (300 nM) caused a 52 ± 2% decrease in fluorescence.

Simultaneous application of the opioids DAMGO (pEC50 7.4, Emax 56%), morphine (pEC50

7.0, Emax 61%) and buprenorphine (pEC50 8.6, Emax 24%) inhibited the FSK response in a dose-dependent manner, while having no effect by themselves. The effects of DAMGO were blocked by pertussis toxin. This assay represents a simple, robust method for real- time observation of AC inhibition by MOR in CHO cells. It represents an appealing alternative to end-point assays that rely on cAMP accumulation and can avoid potential confounds associated with rapid desensitisation of MOR signalling.

Keywords: opioid, membrane potential, high throughput, cAMP, calcitonin

80 INTRODUCTION

Opioid analgesics are the most widely prescribed drugs in the treatment of moderate to severe pain. Despite their powerful analgesic effects, the use of opioids is limited due to the number of associated adverse affects such as respiratory depression, sedation, constipation and nausea, as well as the development of tolerance. Over time, the development of opioid tolerance and physical or psychological dependence may require

10-fold escalations in dose in order to maintain adequate pain relief.1 Both the analgesic and adverse effects of opioid analgesics are mediated via the -opioid receptor (MOR) subtype.2 MOR mediate their effects via downstream mechanisms including inhibition of adenylyl cyclase (AC) activity via G i/o subunits, inhibition or activation of ion channels via G subunits and activation of MAPK signalling via -arrestin.3

One of the hallmarks of MOR receptor activation is the inhibition of AC activity, leading to a decrease in the production of cAMP. Changes in cAMP-dependent signalling are also hallmarks of processes associated with chronic opioid receptor activation.4,5 cAMP is an important cellular second messenger, mediating a diverse range of physiological processes via activation of cAMP-dependent protein kinase A (PKA), exchange protein directly regulated by cAMP (EPAC) as well as directly via cAMP-gated ion channels.6,7 The measurement of cAMP accumulation is frequently used as a sensitive endpoint assay in studies of both acute and chronic MOR signalling. A number of techniques have been developed for quantifying cAMP accumulation in recent years, particularly in the rapidly growing field of high-throughput screening (HTS). AC activity can be measured in a number of ways including measurement of the accumulation of [3H]-cAMP, by [3H]- cAMP binding displacement assays, by reporter gene assays utilizing cAMP-dependent transcription factors, by ELISA assays that measure cAMP-like immunoreactivity, and

81 through measurements of cAMP-dependent protein/protein interactions using FRET.8

These assays often require cell lysis, are usually single time-point, and require the addition of multiple reagents or transfection of reporter constructs. Importantly, many assays of AC activity require significant incubation times to allow appropriate cAMP accumulation.9, 10

In studies of MOR signalling, prolonged incubation times can pose a significant and underappreciated problem as MOR signalling undergoes rapid desensitisation followed by receptor internalisation during agonist exposures as short as 5 - 10 minutes.11 As the incubation periods in most cAMP assays are at least 10 – 20 minutes, during which time opioid exposure is sustained, these assays are likely to be measuring the combined effects of receptor activation, desensitisation, internalisation and even resensitisation.11,12

In this study, we sought to develop a simple, rapid assay of AC modulation. Here we report an assay of MOR inhibition of AC in intact CHO cells using a proprietary membrane potential sensitive dye. The assay is rapid, real-time, robust and requires minimal preparation. This assay should also obviate the problem of MOR receptor desensitisation during baseline measurements of AC inhibition.

MATERIALS AND METHODS

MOR transfection and cell culture

Flp-In T-Rex Chinese hamster ovary (CHO) cells were created as follows. CHO Flp-In cells were grown in minimal essential medium alpha (Invitrogen, Melbourne, Australia) containing 5% fetal bovine serum (FBS) and 100 g/ml zeocin. They were transfected with pcDNA6TR (tet-repressor plasmid) using Fugene 6 reagent (Promega, Alexandria,

Australia) and selected with 10 g/ml blasticidin and 100 g/ml zeocin. Six individual clones were isolated and transiently transfected with the plasmid pcDNA5-FRT-TRPV1,

82 which encodes the Transient Receptor Potential vanilloid 1 receptor (TRPV1) under control of a tetracycline-sensitive repressor. Clones were then tested for successful induction with 1 g/ml tetracycline using a plated-based calcium assay of TRPV1 receptor activation.13 One CHO-FRT-TR cell line was chosen and stably transfected with a pcDNA5 construct encoding the haemagglutinin-tagged human -opioid receptor cDNA together with the pOG44 (Flp recombinase plasmid) using the transfectant Fugene

(Promega). The HA-tagged human -opioid receptor was synthesised by Genscript

(Piscataway, New Jersey, USA). Cells expressing MOR were selected using hygromycin

B (500 µg/mL) and grown to confluency. The selected cells were then cultured in

Dulbecco’s Modified Eagle Medium (DMEM) containing 10% FBS, 100U penicillin/streptomycin and 500 g/mL hygromycin B up to passage 5. Hygromycin concentration was reduced to 200 g/mL beyond passage 5. Cells were passaged at 80% confluency as required. Assays were carried out on cells up to 30 passages. Cells for assays were grown in 75 cm2 flasks and used at greater than 80% confluence. The day before the assay cells were detached from the flask with trypsin/EDTA (Sigma) and resuspended in 10 ml of Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin and 15 mM glucose. hMOR receptor expression was induced with

2 g/mL tetracyclin 20 hrs prior to the assay. The cells were plated in a volume of 90 µl in black walled, clear bottomed 96-well microplates (Corning) and incubated overnight at

o 37 C in ambient CO2.

Membrane potential assay

Membrane potential was measured using a FLIPR Membrane Potential Assay kit (blue) from Molecular Devices. The dye was reconstituted with assay buffer containing (in mM),

NaCl 145, HEPES 22, Na2HPO4 0.338, NaHCO3 4.17, KH2PO4 0.441, MgSO4 0.407,

83 MgCl2 0.493, CaCl2 1.26, glucose 5.56 (pH 7.4, osmolarity 315 ± 5). Prior to the assay, cells were loaded with 90 l/well of the dye solution without removal of the L-15, giving

an initial assay volume of 180 l/well. Plates were then incubated at 37°C at ambient CO2 for 60 minutes. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at

565 nm. Baseline readings were taken every 2 seconds for at least 2 minutes, at which time either drug or vehicle was added in a volume of 20 l. Further additions were made in volumes of 20 µl, as indicated. The background fluorescence of cells without dye or dye without cells was negligible. Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle addition. The final concentration of DMSO was not more than 0.1%, and this concentration did not produce a signal in the assay.

Drugs and Chemicals

Unless otherwise noted, tissue culture reagents and buffer salts were from Invitrogen or

Sigma. [D-Ala2, N-MePhe4, Gly-ol]-enkephalin (DAMGO), was purchased from Auspep

(Tullamarine, Australia). Morphine was a kind gift from the Department of Pharmacology,

University of Sydney. Buprenorphine was from the National Measurement Institute

(Lindfield, Australia). Rp-8-(4-chlorophenylthio)adenosine-3',5'- cyclic mono- phosphorothioate, (Rp-8-CPT-cAMPS), Sp-8-(4-chlorophenylthio)adenosine- 3', 5'- cyclic monophosphorothioate (Sp-8-CPT-cAMPS) and 8-(4-chlorophenylthio)-2'-O- methyladenosine-3',5'-cyclic monophosphate, acetoxymethyl ester (8-CPT-2Me-cAMP) were from Biolog (Bremen, Germany). Membrane potential dye (blue) was from

Molecular Devices (Sunnyvale, Califormia, USA). Calcitonin was from Bachem

(Bubendorf, Switzerland). Nigericin was from Enzo Life Sciences (Farmingdale, NY,

84 USA). Forskolin, naloxone, H89, staurosporine, glibenclamide and 4-aminopyridine were from Ascent Pharmaceuticals (Bristol, UK). 1,9-dideoxyforskolin and tetraethylammonium

(TEA) were from Sigma Aldrich (Castle Hill, Australia). Pertussis toxin (PTX) and VU-

591 were from Tocris Bioscience (Bristol, UK). Charybdotoxin (CHX) was from Alexis

Biochemicals (San Diego, US). KT-5720 was from Cayman Chemicals (Michigan, US).

Data

Unless otherwise noted, data is expressed as mean ± s.e.m. of at least 5 determinations made in duplicate or triplicate. Concentration response curves were fit with a 4 parameter logistic equation using Graphpad Prism (Graphpad). Statistical comparisons were made with an unpaired Student’s T-test, unless otherwise noted. P < 0.05 was considered significant. All channel and receptor nomenclature is consistent with the British Journal of

Pharmacology Guide to Receptors and Channels. 14

To calculate Z’, a measure of the robustness of the assay and indication of its suitability for

HTS, the assay was performed on three separate occasions using assay buffer as the minimum response and either 300 nM FSK or 300 nM FSK with 3 µM DAMGO as the maximum response in 96-well plates. The Z’ factor was calculated as outlined in Zhang et al.15

RESULTS

Hyperpolarisation of CHO cells by calcitonin and forskolin

CHO cells express calcitonin receptors, Gs-coupled G protein-coupled receptors (GPCR) which stimulate AC.16 In CHO-MOR cells loaded with the proprietary membrane potential dye, addition of rat calcitonin produced an immediate decrease in fluorescence, consistent with hyperpolarisation of the cells (Figure 1A). The fluorescent signal continued to slowly

85 decrease for 10 min after the addition of calcitonin, after which fluorescent signals remained stable for the remainder of the assay. The decrease in fluorescence was concentration dependent, with Emax of 46 ± 3% and pEC50 of 9.4 ± 0.1 (n = 5, Figure 1B).

Addition of forskolin (FSK) to CHO cells loaded with membrane potential dye produced a similar decrease in fluorescence to that observed with calcitonin, with the fluorescent signal stabilising 5 min after the addition of FSK (Figure 1A). The decrease in fluorescence for FSK was concentration dependent, with a maximal effect (EMax) of 52 ±

2% and pEC50 of 7.3 ± 0.1 (n = 6, Figure 1B). The AC-inactive FSK analogue 1,9- dideoxyforskolin (1 µM) did not produce a change in fluorescence (n = 5, Figure 3C, P >

0.1).

Opioid inhibition of adenylyl cyclase

Application of opioids alone did not affect the membrane potential of CHO-MOR cells.

However, when the MOR agonist DAMGO was added together with a sub-maximally effective concentration of FSK (300 nM), DAMGO inhibited FSK induced membrane hyperpolarisation in a concentration dependent manner, consistent with inhibition of cAMP generation (Figure 2A). This response was blocked by naloxone (1 µM, Figure 2A).

We measured the effects of opioids on the FSK-induced hyperpolarisation 5 min after co- application of the drugs. DAMGO inhibited the FSK-induced hyperpolarisation with a pEC50 of 7.4 ± 0.1 and a maximum inhibition of 56 ± 3%. Maximum DAMGO inhibition of FSK-induced hyperpolarisation was reduced to 40 ± 2% (p < 0.02) when measured 10 minutes after FSK addition, while the potency was unchanged. Pretreatment of cells overnight with pertussis toxin (200 ng/ml) significantly reduced the inhibition of the FSK response by DAMGO (1 µM), inhibition was 81 ± 12% in control, and 11 ± 0.5 % after

PTX treatment (P < 0.01, n=3). The change in fluorescence produced by FSK was unaffected by PTX treatment (46 ± 2% in control; 45 ± 5% after PTX, P = 0.73). FSK-

86 induced hyperpolarisation was observed in CHO-MOR cells where MOR expression had not been induced by tetracycline, however the hyperpolarisation was not inhibited by opioids (data not shown). DAMGO (1 µM) also reduced membrane hyperpolarisation produced by 10 nM calcitonin from 40 ± 1.6% to 23 ± 1.2% (P < 0.01).

Figure 1: Stimulating adenylyl cyclase hyperpolarises CHO cells. The membrane potential of CHO cells was determined as outlined in the Methods. Both calcitonin and forskolin hyperpolarise CHO-MOR cells in a concentration dependent manner. A) Representative traces showing the raw fluorescence (RFU) from individual wells of a 96- well plate. Calcitonin (10 nM) or forskolin (1 µM) were added to CHO-MOR cells for the duration of the bar. B) Concentration response curves illustrating the effects of calcitonin and forskolin on the membrane potential of CHO-MOR cells. The calcitonin Emax was 46 ±

87 3% and pEC50 was 9.4 ± 0.1. Forskolin Emax 52 ± 2% and pEC50 was 7.3 ± 0.1. Each point represents the mean ± s.e.m. of at least 5 experiments performed in triplicate.

Figure 2: Opioids inhibit forskolin-stimulated membrane hyperpolarisation in CHO cells. The membrane potential of CHO-MOR cells was determined as outlined in the Methods. A) Example traces showing the effects of forskolin (FSK, 300 nM), FSK and DAMGO (1 µM) applied simultaneously and FSK and DAMGO added in the presence of naloxone (NAL, 1 µM) which had been added 10 minutes earlier. FSK and DAMGO were added for the duration of the bar. B) Concentration-response curves for morphine, buprenorphine and DAMGO inhibition of the hyperpolarisation produced by FSK (300 nM). DAMGO inhibition of FSK stimulated AC activation was concentration-dependent, with an Emax of 56 ± 3 %, and pEC50 of 7.4 ± 0.1, morphine had an Emax of 61 ± 7% and pEC50 of 7.0 ± 0.2, Buprenorphine showed partial agonist activity with Emax of 24 ± 4% and pEC50 of 8.6 ± 0.5. C) Preincubation of the cells with increasing concentrations of naloxone (3, 10, 100 nM) produced a parallel shift in the concentration-response curve of morphine, with a pA2 of -8.5 ± 0.1, (2.9 ± 0.5 nM, n=4).

88 A RFU (% Predrug)

100 FSK + DAMGO (1 µM)

75 FSK (300 nM) FSK+DAMGO in naloxone (1 µM) 50

25

0 0 120 240 360 480 600 720 Time (s) B Inhibition of FSK Response ( % Control) DAMGO, EC 25 nM 60 50 Morphine, EC50 70 nM

Bup, EC50 3 nM 40

20

0 -10 -9 -8 -7 -6 -5 [Opioid] log M C Inhibition of FSK Response ( % Control)

60 [naloxone] 0 3 nM 40 10 nM 100 nM

20

0 -9 -8 -7 -6 -5 [Morphine] log M

89 We assessed the capacity of this assay to reliably detect agonists of differing efficacy by examining the effects of morphine and buprenorphine (Figure 2B). Morphine and buprenorphine have previously been shown to have partial agonist activity at MOR.17,18

Both agonists inhibited FSK-stimulated AC activation. Morphine had a similar efficacy as

DAMGO, with Emax of 61 ± 7%, and pEC50 of 7.0 ± 0.2. Buprenorphine showed lower efficacy with Emax of 24 ± 4%, and pEC50 of 8.6 ± 0.5. Addition of increasing concentrations of naloxone produced a parallel shift in the concentration response curve for morphine, with a pA2 of -8.5 ± 0.1 (2.9 ± 0.5 nM, n=3), a value consistent with the reported binding affinity of naloxone at human MOR19 (Figure 2C).

Mechanism of forskolin induced hyperpolarisation of CHO-MOR cells

We sought to determine the mechanism by which AC activation caused membrane hyperpolarisation in CHO-MOR cells. The increase in cAMP resulting from AC activation can lead to activation of PKA, EPAC, or cAMP-gated ion channels.6-7, 20 The membrane permeable cAMP analogue Sp-8-CPT-cAMPs (100 µM), a direct activator of PKA, mimicked the FSK response, producing a 44 ± 4% decrease in fluorescence (n=5, Figure

3A). Rp-8-CPT-cAMPs, a cAMP analogue that inhibits activation of PKA, did not produce a change in membrane fluorescence. The hyperpolarisation produced by Sp-8-

CPT-cAMPs was not affected by DAMGO (Figure 3B), consistent with Sp-8-CPT-cAMPs producing its effects downstream of AC, potentially by acting directly on PKA. However, moderate concentrations of PKA inhibitors H89 (100 nM - 1 µM), KT5270 (100 nM – 1

µM) and staurosporine (1 µM) did not affect the hyperpolarisation produced by forskolin, and at higher concentrations (10 µM and above) the protein kinase inhibitors produced substantial hyperpolarisations by themselves. The cAMP analog 8-CPT-2Me-cAMP (100

µM), which selectively activates EPAC and not PKA, did not significantly affect cellular fluorescence (n=4, Figure 3C, P > 0.5).

90 Membrane hyperpolarisation usually occurs by efflux of K ions from cells. The non- specific K channel blockers tetraethyl ammonium chloride (TEA, up to 10 mM) or 4 aminopyridine (4-AP, up to 1 mM) did not inhibit the FSK-induced hyperpolarisation

(Figure 4B). Additionally, the hyperpolarisation produced by FSK was not affected by the more selective KATP channel blocker glibenclamide (10 M), the renal outer medullary

potassium channel (Kir 1.1) blocker VU-591 (10 M) or charybdotoxin (10 M), a blocker of high-conductance Ca activated K channels (KCa1.1). To determine if FSK-induced

membrane hyperpolarisation was due to K efflux, the extracellular K concentration ([K]Ex) was increased to 30 mM and 75 mM in order to make the reversal potential (Ke) for K less negative. This will reduce K efflux and associated membrane potential changes when K channels are opened. [K]Ex was adjusted by dissolving the membrane potential dye in

HBSS where NaCl was substituted by KCl. The effects of a maximally effective concentration of FSK (10 µM) was reduced by about 50% in 30 mM [K]Ex, and almost completely when [K]Ex was 75 mM (Figure 4A), suggesting FSK-induced membrane hyperpolarisation is mediated by efflux of K from cells. The hyperpolarisations produced by 100 nM, 1 µM and 10 µM FSK in the 3 concentrations of [K]Ex were analysed by 2 way

ANOVA and found a main effect of FSK (P < 0.003) and [K]Ex (P < 0.0001). Subsequent analysis of the effect of [K]Ex on each concentration of FSK using Tukey’s multiple comparison test indicated that the hyperpolarisation produced by each concentration of

FSK differed for each concentration of [K]Ex (P < 0.05 for each). The changes in fluorescence produced by altering the membrane K permeability were independently assessed by incubating CHO-MOR cells with nigericin, a potassium selective antibiotic ionophore.21 A maximally effective concentration of nigericin (1 µM) produced a decrease in fluorescence signal of 70 ± 3%, which was not decreased any further upon the addition

91 of 10 µM FSK (Figure 4C). The fluorescent signal observed after nigericin incubation may

21 reflect the signal when the membrane potential of the cells is driven to EK.

Figure 3: Protein kinase A activators mimic the effects of forskolin. The membrane potential of CHO-MOR cells was determined as outlined in the Methods. A) The cAMP analog Sp-8-CPT-cAMPs (100 M, black trace), but not Rp-8-CPT-cAMPs (blue trace) mimicked the effect of FSK (100 nM, red trace). B) The hyperpolarisation produced by 30 M Sp-8-CPT-cAMPs (black trace) was not inhibited by the simultaneous addition of 1

M DAMGO (blue trace). C) The AC-inactive FSK analogue 1,9-ddFSK (1 M) and the EPAC selective cAMP analogue 8-CPT-2Me-cAMP (100 µM) had little effect on CHO- MOR membrane potential when compared with 1 M FSK. Bars represent the mean ± s.e.m of 4-5 determinations in triplicate.

92 A RFU (% predrug) 100 Rp-cAMPS (100 µM) Sp-8-CPT-cAMPS (30 µM) 75 FSK (100 nM) 50

25

0 0 120 240 360 480 600 720 Time (s) B RFU 1400

1200

1000

800 Sp-8-CPT-cAMPS (30 µM)

600 + DAMGO (1 µM) 400

200

0 0 120 240 360 480 600 720 Time (s) C D Fluorescence (% Predrug) 60

50

40

30

20

10

0 Forskolin 1,9-ddFSK 8-CPT-2Me-cAMP (1 µM) (1 µM) (100 µM)

93

Figure 4: The forskolin-induced hyperpolarisation is dependent on extracellular K concentration. The membrane potential of CHO-MOR cells was determined as outlined in the Methods. A) The effects of forskolin (FSK) were reduced as the extracellular K concentration was changed from 5 mM to 30 mM or 75 mM. K was replaced with equimolar Na. The hyperpolarisations produced by 100 nM, 1 µM and 10 µM FSK in the

3 concentration of [K]Ex were analysed by 2 Way ANOVA and found a main effect of FSK

(P < 0.003) and [K]Ex (P < 0.0001). B) Membrane hyperpolarisation produced by FSK (300 nM) was not affected the non-specific K+ channel blockers tetraethylammonium chloride (TEA) or 4-aminopyridine (4-AP), the selective KATP channel blocker glibenclamide (Glib), charybdotoxin (CHTX), a blocker of high-conductance Ca activated K channels and the renal outer medullary K channel channel blocker VU-591. C) The K selective ionophore nigericin (1 µM) produced a decrease in fluorescence that occluded the effects of a high concentration of FSK (10 µM).

94

95 We assessed the suitability of this assay for HTS by calculating the Z’ factor, a measure of the assay robustness. An assay with a Z’ factor of between 0.5 and 1 is an appropriate assay in terms of signal-to-noise ratio and data reproducibility.15 The Z’ factor for this assay was calculated for both the 300 nM FSK response and for inhibition of the FSK response with 3 µM DAMGO in multiple experiments. The Z’ values for FSK alone were

0.7, 0.7, and 0.8, and for FSK + DAMGO were 0.7, 0.7 and 0.7, indicating that the assay is suitable for HTS.

DISCUSSION

In this study, we have developed a real time, no wash, fluorescence-based assay for MOR- mediated inhibition of AC in intact CHO-K1 cells. We used a proprietary membrane potential sensitive dye from Molecular Devices to measure membrane hyperpolarisation following FSK stimulated AC activation.22 Activation of endogenous calcitonin receptors in CHO-K1 cells16 produced a reduction in fluorescence similar to that seen following FSK application, and in both cases this reduction was less than that produced by application of the K-selective ionophore nigericin21. The fluorescent signal rapidly and reliably decreased after FSK application, and this decrease in signal was inhibited by the simultaneous application of opioid agonists.

CHO cells are frequently used in assays of opioid inhibition of AC activity. The elevation of cAMP in CHO cells following FSK application, as well as the ability of opioids to inhibit this elevation of cAMP is well documented.16, 23-24 In this assay, opioid inhibition of

FSK-stimulated membrane hyperpolarisation is consistent with opioid inhibition of AC, one of the hallmarks of MOR activation.3 CHO cells have been shown to express AC subtypes VI and VII.25 Opioid modulation of AC activity is isozyme specific, with acute opioid treatment shown to inhibit Gs-stimulated AC-VI activity and potentiating Gs-

96 stimulated AC-VII.26 Because AC-VII has been reported to be insensitive to FSK,32 we chose to use FSK rather than calcitonin as the AC-stimulating agent, as this should avoid the confounding effects of opposite MOR modulation of the Gs-mediated stimulation of

AC-VI and AC-VII.

Morphine and buprenorphine are lower efficacy agonists at MOR17,18, however the measured efficacy of a compound depends on the assay being used. In the present assay, morphine acted as a full agonist, with an Emax similar to that of DAMGO. This likely reflects the relatively low receptor occupancy required for AC inhibition, particularly as we were stimulating AC with a submaximally effective concentration of forskolin. Many previous studies have reported morphine to be a full agonist in assays of AC inhibition.4,28

The measurement of MOR inhibition of AC is often performed using end-point assay techniques with lengthy inhibition times. As shown in our assay, some sensitivity would be lost using this approach. Inhibition of the FSK-stimulated membrane hyperpolarisation peaked at approximately 5 minutes after the addition of FSK and opioid. After this time, the fluorescent signal gradually decreased further. A single time-point measurement of cAMP accumulation after 10 minutes in our assay shows a reduction in the efficacy of

DAMGO in AC inhibition, with Emax decreased from 56% to 40%, possibly reflecting receptor desensitisation. While we chose to measure at a time point when the FSK- induced hyperpolarisation signal had plateaued, it would be a simple matter to measure at any time point after addition of the drugs, enabling the virtually instantaneous measurement of cell responses after the addition of FSK and/or opioid. Depending on the agonist employed, MOR rapidly desensitises and/or internalises, with up to 50% of MOR internalized within 5 minutes of agonist exposure in some cells.11-12,20 AC assays using single time point measurement of cAMP accumulation after incubation times of up to 20

97 minutes are measuring the summed output of signalling from activated, desensitized and internalized MOR, and thus give little insight into the real-time dynamics of the effects of

MOR activation on cAMP mediated signalling.

The measurement of real-time AC inhibition has previously only been achievable by the use of complex techniques such as transfection of reporter genes or proteins with cAMP binding domains, or bioluminescence resonance energy transfer (BRET) assays.8, 29-32

These assays can be very useful for studying cAMP signalling, although each has its potential weaknesses as well as strengths.8 Bioluminescence assays such as the Glosensor

(Promega) can provide a continuous reading of cAMP levels, allowing kinetic studies and repeated drug applications30. However, this assay requires transfection of a luciferase sensor construct with cAMP binding domains and use of specialized reagents, in addition to any stable transfections of the receptors of interest. For the kind of studies described in this paper – inhibition of cAMP accumulation by a Gi/Go-coupled GPCR, the developers of the Glosensor assay recommend a 5-10 minute pre-incubation with agonist before the addition of FSK, which is likely to be unsuitable for studies of rapidly desensitizing GPCR such as MOR.30 The assay described here is simpler than these assays, requiring transfection of the receptor of interest only and addition of a single assay reagent. The drug responses can be observed immediately following agonist addition. The Z’ calculated for this assay were similar to those reported for the Glosensor assay.31,32 However, as the plate reader format means that we are unable to study recovery of AC activity following wash of agonists, and the kinetic and stoichiometric relationship between the FSK-induced rise in AC levels and hyperpolarisation of the CHO cells is unknown.

Activation of the endogenous Gs-coupled CTR receptors in CHO cells caused membrane hyperpolarisation similar to that produced by the AC activator FSK. Membrane

98 hyperpolarisation in CHO cells by AC activation has not been shown previously, so we sought to determine the mechanism by which membrane hyperpolarisation occurs.

Endogenous K channels are not well defined in CHO cells, and CHO cells are frequently used as heterologous expression systems for recombinant K channels due to the apparent low levels of native K channel activity.33 The non-specific K channel inhibitors TEA and 4-

AP did not inhibit FSK induced membrane hyperpolarisation, nor did the more specific K channel inhibitors glibenclamide, VU-591 or charybdotoxin. However, when the reversal potential for K was made less negative by increasing [K]Ex, membrane hyperpolarisation was reduced, and hyperpolarisation was essentially abolished when [K]Ex was increased to

75 mM. Furthermore, the FSK induced membrane hyperpolarisation was mimicked and occluded by the addition of the K-selective ionophore nigericin. This suggests that FSK induced membrane hyperpolarisation is due to the movement of K ions through native K channels in CHO cells.

Membrane hyperpolarisation resulting from application of FSK and calcitonin is due to the activation of AC.16,22 The resulting elevation of cAMP levels in the cell leads to the activation of PKA, as well as EPACs. The cAMP analogue Sp-8-CPT-cAMPs, a direct activator of PKA, mimicked the effects of FSK, while Rp-8-CPT-cAMPs, an inhibitor of cAMP activation of PKA, was inactive. 8-CPT-2Me-cAMP, a selective activator of

EPAC, also produced no effect. Together these data are consistent with the idea that membrane hyperpolarisation is occurring via PKA activation. However, we were unable to inhibit the effects of FSK with modest concentrations of protein kinase inhibitors, and higher concentrations of these drugs themselves hyperpolarized the CHO cells. Novel cAMP-dependent signal transduction pathways are still being discovered,34 and it may be that the observed hyperpolarisation of CHO cells is mediated by such a mechanism.

99 Assays of AC inhibition represent one of the most straightforward ways of studying G (as opposed to G ) signalling in a high throughput environment. The assay described here offers a novel approach for measuring MOR-mediated AC inhibition in intact CHO cells, and has the advantages of being both real-time and reflecting the naturalistic coupling of

MOR to the signalling pathway. The lack of a defined mechanism for the hyperpolarisation does not detract from the utility of the assay for acute studies, however, the assay may not be suitable for studies of more complex signalling cascades such as those potentially underlying agonist-induced receptor regulation. Nevertheless, our results show the membrane potential assay to be a rapid, reliable and inexpensive method for assessing opioid activation of MOR in CHO cells, and may be scaled up to enable high- throughput screening. Coupled with our recent description of a HTS-appropriate membrane potential assay of G signalling in AtT-20 cells35, it is clear that multiple aspects of GPCR signalling can be studied in a relatively simple, non-invasive and efficient manner using simple reagents which report changes in basic cellular properties such as membrane potential.

Acknowledgements

This work was supported by NHMRC Grant 1011979 to MC. AK was supported by a

MQRes Postgraduate Scholarship from Macquarie University.

100 References

1. Buntin-Mushock C.; Phillip L.; Moriyama K. et al. Age-dependent opioid

escalation in chronic pain patients. Anesth. Analg. 2005, 100, 1740-1745.

2. Matthes H.W.D.; Maldonado R.; Simonin F. et al. Loss of morphine-induced

analgesia, reward effect and withdrawal symptoms in mice lacking the -opioid-

receptor gene. Nature 1996, 383, 819-823.

3. Connor M.; Christie M.J. Opioid receptor signalling mechanisms. Clin. Exp.

Pharmacol. Physiol. 1999, 26, 493-499.

4. Avidor-Reiss T.; Bayewitch M.; Levy R.; et al. Adenylylcyclase supersensitisation

in µ-opioid receptor transfected Chinese hamster Ovary cells following chronic

opioid treatment. J. Biol Chem. 1995, 270, 29732-29738.

5. Christie M.J. Cellular neuroadaptations to chronic opioids: tolerance, withdrawal

and addiction. Brit. J. Pharmacol. 2008, 154, 384-396.

6. Biel M.; Michalakis S. Cyclic nucleotide-gated channels. Hand. Exper. Pharmacol.

2009, 191, 111-136.

7. Borland G.; Smith B.O.; Yarwood S.J. EPAC proteins transduce diverse cellular

actions of cAMP. Brit. J. Pharmacol. 2009, 158, 70-86.

101 8. Hill S.J.; Williams C.; May L.T. Insights into GPCR pharmacology from the

measurement of changes in intracellular cyclic AMP; advantages and pitfalls of

differing methodologies. Brit. J. Pharmacol. 2010, 161, 1266-1275.

9. Gabriel D.; Vernier M.; Pfeifer M.J. et al. High throughput screening technologies

for direct cyclic AMP measurement. Assay Drug Dev. Techn. 2003, 1, 291-303.

10. Williams C. cAMP detection methods in HTS: selecting the best from the rest. Nat.

Rev. Drug. Discov. 2004, 3, 125-135.

11. Connor M.; Osborne P.B.; Christie M.J. µ-Opioid receptor desensitisation: Is

morphine different? Brit. J. Pharmacol. 2004, 143, 685-696.

12. Dang V.C.; Christie M.J. Mechanisms of rapid opioid receptor desensitisation,

resensitisation and tolerance in brain neurons. Brit. J. Pharmacol. 2012, 165, 1704-

1716.

13. Veldhuis N.A.; Lew M.J.; Abogadie F.C. et al. N-Glycosylation determines ionic

permeability and desensitisation of the TRPV1 capsaicin receptor. J. Biol. Chem.

2012, 287, 21765-21772.

14. Alexander S.P.H.; Mathie A.; Peters J.A. Guide to receptors and channels (GRAC),

5th edn. Brit. J. Pharmacol. 2011, 164 (Suppl.1), S1-S324.

102 15. Zhang, J.; Chung, T.D.Y.; Oldenburg, K.R. A Simple Statistical Parameter for use

in Evaluation and Validation of High Throughput Screening Assays. J. Biomol.

Screen. 1999, 4, 67-73.

16. Wang Y.; Kong Y.; Shei G.J. et al. Development of a cyclic adenosine

monophosphate assay for Gi-coupled G protein-coupled receptors by utilizing the

endogenous calcitonin activity in Chinese hamster ovary cells. Assay Drug Dev.

Techn. 2011, 9, 522-531.

17. Borgland S.L.; Connor M.; Osborne P.B. et al. Opioid agonists have different

efficacy profiles for G protein activation, rapid desensitisation, and endocytosis of

mu-opioid receptors. J. Biol. Chem. 2003, 278, 18776-18784.

18. Virk M.S.; Arttamangkul S.; Birdsong W.T. Buprenorphine is a weak partial

agonist that inhibits opioid receptor desensitisation. J. Neurosci. 2009, 29, 7341-

7348.

19. Brillet K.; Kieffer B.L.; Massotte D. Enhanced spontaneous activity of the mu

opioid receptor by cysteine mutations: characterisation of a tool for inverse agonist

screening. BMC Pharmacol. 2003, 3,14.

20. Skalhegg B.S.; Tasken K. Specificity in the cAMP/PKA signalling pathway:

Differential expression, regulation, and subcellular localisation of subunits of PKA.

Front. Biosci. 1997, 5, D678 - D693.

103 21. Doebler J.A. Comparative effects of carboxylic ionophores on membrane potential

and resistance of NG108-15 cells. Toxicol. in Vitro 2000, 14, 235-243.

22. Seamon K.B.; Padgett W.; Daly J.W. Forskolin: Unique diterpene activator of

adenylate cyclase in membranes and in intact cells. Proc. Natl. Acad. Sci. USA

1981, 78, 3363-3367.

23. Bunzow J.R.; Zhang G.; Bouvier C. et al. Characterisation and distribution of a

cloned rat -opioid receptor. J. Neurochem. 1995, 64, 14-24.

24. Chakrabarti S.; Prather P.L.; Yu L. et al. Expression of the µ-opioid receptor in

CHO cells Ability of µ-opioid ligands to promote alpha-azidoanilido [P-32] GTP

labeling of multiple G-protein alpha subunits. J. Neurochem. 1995, 64, 2534-2543.

25. Varga E.V.; Stropova D.; Rubenzik M. et al. Identification of adenylyl cyclase

isoenzymes in CHO and B82 cells. Eur. J. Pharmacol. 1998 348, R1-R2.

26. Avidor-Reiss T.; Nevo I.; Saya D. et al. Opiate-induced adenylyl cyclase

superactivation as isozyme-specific. J. Biol. Chem. 1997, 272, 5040-5047.

27. Barak L.S.; Salahpour A.; Zhang X. et al. Pharmacological characterisation of

membrane-expressed human trace amine-associated receptor 1 (TAAR1) by a

bioluminescence resonance energy transfer cAMP Biosensor. Mol. Pharmacol.

2008, 74, 585-594.

104 28. Nickolls S.A.; Waterfield A.; Williams R.E. et al. Understanding the effect of

different assay formats on agonist parameters: A study using the µ-opioid receptor.

J. Biomol. Screen. 2011, 16, 706-716.

29. Jiang L.I.; Collins J.; Davis R. et al. Use of a cAMP BRET sensor to characterize a

novel regulation of cAMP by the sphingosine 1-phosphate/G13 pathway. J. Biol.

Chem. 2007, 282, 10576-10584.

30. Binkowski B.F.; Fan F.; Wood K.V. Luminescent biosensors for real time

monitoring of intracellular cAMP. Methods Mol. Biol. 2011, 756, 263-271.

31. Pantel J.; Williams S.Y.; Mi D. et al. Development of a high throughput screen for

allosteric modulators of melanocortin-4 receptor signalling using a real time cAMP

assay. Eur. J. Pharm. 2011, 660, 139-147.

32. Binkowski B.F.; Butler B.L.; Stecha P.F. et al. A luminescent biosensor with

increased dynamic range for intracellular cAMP. ACS Chem. Biol. 2011, 6, 1193-

1197.

33. Yan L.; Herrington J.; Goldberg E. et al. Stichodactyla helianthus peptide, a

pharmacological tool for studying Kv3.2 Channels. Mol. Pharmacol. 2005, 67,

1513-1521.

34. Emery A.C.; Eiden L.E. Signalling through the GPCR PAC induces

neuritogenesis via a single linear cAMP- and ERK-dependent pathway using a

novel cAMP sensor. FASEB J. 2012, 26, 3199-3211.

105 35. Knapman A.; Santiago M.; Du Y.P. et al. A continuous, fluorescence-based assay

of µ-opioid receptor activation in AtT-20 cells. J. Biomol. Screen. 2013, 18, 269-

276.

106 Chapter 4

A continuous, fluorescence-based assay of μ-opioid receptor activation in AtT-20 cells

Alisa Knapman, Marina Santiago, Yan Ping Du, Philip R. Bennallack,

Macdonald J. Christie and Mark Connor

This paper was published in 2013 in the Journal of Biomolecular Screening 18: 269-276.

Yan Ping Du and Macdonald Christie constructed AtT-20 cells stably expressing mouse

MOPr, and quantified receptor expression. Marina Santiago and Phillip Bennallack designed and performed desensitisation experiments. Mark Connor assisted in experimental design and manuscript preparation. All other work is my own.

107 ABSTRACT

Opioids are widely prescribed analgesics, however their use is limited due to development of tolerance and addiction, as well as high variability in individual response. The development of improved opioid analgesics requires high-throughput functional assays to assess large numbers of potential opioid ligands. In this study, we assessed the ability of a proprietary “no wash” fluorescent membrane potential dye to act as a reporter of µ-opioid receptor (MOR) activation and desensitisation via activation of G protein-coupled inwardly rectifying potassium channels. AtT-20 cells stably expressing mouse MOR were assayed in 96-well plates using the Molecular Devices FLIPR membrane potential dye.

Dye emission intensity decreases upon membrane hyperpolarisation. Fluorescence decreased in a concentration-dependent manner upon application of range of opioid ligands to the cells, with high efficacy agonists producing a decrease of 35% to 40% in total fluorescence. The maximum effect of morphine faded in the continued presence of agonist, reflecting receptor desensitisation. The effects of opioids were prevented by prior treatment with pertussis toxin and blocked by naloxone. We have demonstrated this assay to be an effective method for assessing ligand signalling at MOR which may potentially be scaled up as an additional HTS technique for characterizing novel opioid ligands.

108 INTRODUCTION

Opioid analgesics are the most widely prescribed drugs in the treatment of moderate to severe pain. Despite their powerful analgesic effects, the use of opioids in the treatment of chronic pain can be problematic due to the development of tolerance and physical or psychological dependence1. Over time, increasing doses of opioids become necessary to maintain analgesia. The increased toxic side effects such as respiratory depression, constipation and nausea associated with escalating doses of opioids can reach unacceptable levels, leading to inadequate pain relief or overdose. Opioids, however, continue to be the mainstay of chronic pain treatment due to a lack of suitable alternative drugs. As such, there is a substantial need for the development of new opioid analgesics, with reduced adverse effects and a decreased ability to produce tolerance. The µ opioid receptor (MOR) is the primary site of action for most clinically important opioid drugs, and as such is the major target for the development of improved opioid analgesic drugs2. It is generally accepted that there are not functionally important subtypes of MOR3, and so aside from novel formulations or delivery strategies, the development of new pharmacotherapies targeting MOR is likely to focus on subtleties of receptor signalling and regulation.

MOR is a member of the G protein-coupled receptor (GPCR) superfamily, ubiquitous cell- surface receptors that act as cellular switches to regulate most cellular signalling processes4. GPCR agonists stabilise active conformations of their receptors, leading to signalling via both the α and βγ subunits of the associated complex and sometimes also via non-G protein-dependent pathways5. GPCR signalling is complex, with different ligands preferentially activating (or inhibiting) distinct signalling pathways at the same receptor5. There is also increasing evidence that distinct ligands also differentially engage pathways which regulate receptor activity during prolonged agonist exposure, a topic of intense investigation with regard to opioid receptors6,7,8. These ideas

109 have lead to renewed interest in the possibility of developing ligands targeting MOR which engage only subsets of signalling systems or regulatory pathways, potentially leading to drugs with more favourable clinical profiles.

Drug development typically involves the screening of large libraries of lead compounds to identify those capable of binding to and signalling via a receptor. The vast array of lead compounds available requires high-throughput screening methods for efficient detection of potential therapeutic compounds. Radioligand binding studies are often used to identify candidates, but determination of ligand efficacy requires some sort of signalling response.

For opioid receptors, this can be achieved using cell lines expressing engineered G proteins

9 which couple to processes such as intracellular calcium ([Ca]i) mobilisation , cAMP- dependent gene transcription or more traditional assays of adenylyl cyclase (AC) activity which require harvesting or lysing of cells10. In this study we sought to develop a minimally invasive assay that reflected a naturalistic coupling of MOR and which could potentially be used to readily examine agonist regulation of receptor signalling. In mouse pituitary AtT-20 cells, heterologously expressed MOR inhibit native calcium channels

11 12 (ICa) and activate endogenous G protein gated inwardly rectifying potassium channels

(GIRKs), and in both cases this signalling is subject to rapid, agonist-induced regulation.

We assessed the suitability of a proprietary fluorescent membrane potential assay to act as a robust reporter of MOR activation and desensitisation in AtT-20 cells, with the view to potentially providing an assay for high throughput screening of both these important aspects of MOR function.

110 MATERIALS AND METHODS

FLAG-MOR Transfection and Cell Culture

Mouse AtT-20 neuroblastoma cells were stably transfected with the cDNA encoding the

FLAG epitope-tagged mouse μ-opioid receptor using the transfectant Lipofectamine

(Gibco BRL) as previously described11. The pcDNA3 FLAG-MOR construct was a kind gift from Dr. Lakshmi Devi (Mt Sinai School of Medicine, New York, USA). Geneticin

(500 μg/ml) was added to select for clones expressing FLAG-MOR protein. During in situ identification of positive clones using Alexa-488 coupled FLAG-MOR antibody (Sigma,

F7425), 48 potentially suitable single cells were transferred to single wells using a micropipette and were grown to confluence for subsequent determination of cell surface

MOR binding density. MOR binding density was determined on intact cells by incubation with increasing triplicate concentrations of [3H] DAMGO (0.125 - 32 nM; Perkin Elmer,

USA) at 4oC in 50mM Tris-Cl, pH7.4 for 2h. Briefly, approximately 1 x 105 cells were plated in 24-well plate coated with poly-L-lysine overnight. Cells were then rinsed gently twice with 50mM Tris-Cl, pH7.4, placed on ice and incubations in the radioligand were commenced. Non-specific binding (less than 2% of total binding at [3H] DAMGO 5 nM) was determined in the presence of unlabelled DAMGO (10 µM). At the end of the incubation plated cells were rinsed three times with 1 ml 50mM Tris-Cl, pH 7.4 at 4oC.

Cells in each well were then digested for 1 h at room temperature with 100 µl of 1N

NaOH. 100 µl 1N HCl was then added to each well and collected into scintillation vials and bound ligand determined using a liquid scintillation counter (Packard Tricarb, USA).

Specific binding was plotted, and Kd and Bmax for each clone determined using GraphPad

Prism. One clone expressing a moderate density of surface FLAG-MOR was selected for subsequent experiments. The Kd for [3H]-DAMGO binding was 1.2 nM and receptor density was 10.2 pmol/mg protein. After counting cell numbers used for protein determination it was estimated that 2.5 x 107 cells yielded one mg protein. Therefore 10.2

111 pmol/mg protein represents approximately 2.5 x 105 receptors per cell. The selected clone of AtT-20 cells stably expressing mouse FLAG-MOR was then cultured in Dulbecco’s

Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS), 100U penicillin/streptomycin and 300μg/mL G418. Cells were passaged at 80% confluency as required. Assays were carried out on cells up to 25 passages. Cells for assays were grown in 75 cm2 flasks and used at 90% confluence. The day before the assay cells were detached from the flask with trypsin/EDTA (Sigma) and resuspended in 10 ml of

Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin and 15 mM glucose. The cells were plated in volume of 90 µl in black walled, clear bottomed 96- well microplates (Corning) and incubated overnight in ambient CO2.

Membrane Potential Assay

Membrane potential was measured using a FLIPR Membrane Potential Assay kit (blue) from Molecular Devices. The dye was reconstituted with assay buffer supplied with the kit or with a low-K modification. The standard assay buffer contained (mM), NaCl 145,

HEPES 22, Na2HPO4 0.338, NaHCO3 4.17, KCl 5.33, KH2PO4 0.441, MgSO4 0.407,

MgCl2 0.493, CaCl2 1.26, Glucose 5.56 (pH 7.4, osmolarity 315 ± 5). The modified buffer was formulated without the addition of 5.33 mM KCl. Taking into account the K concentration of L-15, the final in-well concentrations of K were 5.55 mM (standard) and

2.88 mM (low K) respectively. Prior to the assay, cells were loaded with 90 μl/well of the dye solution without removal of the L-15, giving an initial assay volume of 180 μl/well.

Plates were then incubated at 37°C at ambient CO2 for 45 minutes. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at 565 nm. Baseline readings were taken every 2 seconds for at least 2 minutes, at which time either drug or vehicle was added in a volume of 20 μl. Further additions were made in volumes of 20 µl, as indicated. The

112 background fluorescence of cells without dye or dye without cells was negligible.

Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle addition, which was less than 2 % for drugs dissolved in assay buffer or DMSO. The final concentration of DMSO was not more than

0.1%.

Data Analysis and Calculation of Z' values

Concentration response data was analysed using PRISM (GraphPad Software Inc., San

Diego, CA), using four-parameter non-linear regression to fit concentration-response curves. The time course of morphine-induced receptor desensitisation was fit with a single phase exponential to obtain an estimated t/2 for the process. To calculate Z', a measure of the robustness of the assay and its suitability for HTS, the membrane potential assay was performed on 3 separate occasions using assay buffer as the negative control and 300 nM

DAMGO as the positive control in 96-well plates. The Z' factor was calculated as outlined in Zhang et al., (1999)13.

RESULTS

In AtT-20 cells loaded with the proprietary membrane potential dye, addition of the peptidergic MOR agonist DAMGO or the prototypic alkaloid opioid morphine produced a rapid decrease in fluorescence, consistent with hyperpolarisation of the cells (Figure 1). It took about 30s for high concentrations of morphine to maximally hyperpolarise the cells

(Table 1). The decrease in fluorescence produced by morphine was concentration dependent and reversed by addition of the opioid naloxone (Figure

2A). Addition of increasing concentrations of naloxone produced a parallel shift in the concentration response curve for morphine, with a pA2 of -8.5 ± 0.1, (2.9 ± 0.5 nM, n=3), a value consistent with the reported binding affinity of naloxone at mouse MOR (2 nM)14

113 (Figure 2B). Pretreatment of cells overnight with pertussis toxin (200 ng/ml) prevented the decrease in fluorescence by DAMGO and morphine (Figure 3). Addition of morphine or

DAMGO to AtT-20 cells not transfected with MOR produced no change in fluorescence.

700 RFU 600

500

400

300 nM DAMGO 300 30 nM DAMGO 3 nM DAMGO 200

100

0 0 60 120 180 240 300 Time (secs)

Figure 1 Figure 1: Example traces of fluorescent signal in the membrane potential assay over 300 sec. Baseline readings were taken every 2 sec for 120 sec, at which point increasing concentrations of DAMGO (3 nM, 30 nM, 300 nM) were added to AtT-20MOR cells, resulting in concentration-dependent decreases in fluorescent signal as shown.

114 Using the assay buffer supplied with the kit, the maximal decrease in fluorescence produced by DAMGO was 31 ± 1% (n=6). We sought to optimise the assay by decreasing extracellular [K]Ex from 5.6 mM to 2.9 mM in order to make the reversal potential (Ke) for

K more negative. The reduction in [K]Ex was achieved by dissolving the dye in saline containing no added KCl. The reduction of [K]Ex resulted in maximally effective concentrations of DAMGO producing a 7 % greater decrease in total fluorescence (P <

0.01, Students T-test). The Emax and pEC50 for DAMGO in 5.6 mM [K]Ex were 31 ± 1% and 8.3 ± 0.1 respectively, while in 2.9 mM [K]Ex, Emax and pEC50 for DAMGO were 38 ±

2% and 8.3 ± 0.1 respectively (Figure 4).

The maximum response elicited by DAMGO was similar to that elicited by somatostatin

(Emax = 39 ± 3.2%, pEC50 8.7), which acts at endogenous sst2 and sst5 receptors to activate

GIRK channels in AtT-20 cells15. We independently assessed the changes in fluorescence produced by altering the membrane K+ permeability by incubating AtT-20 cells with nigericin, a potassium selective antibiotic ionophore16. A maximally effective concentration of nigericin (1 µM) produced a decrease in fluorescence signal of 50 ± 3%, which was not decreased any further upon the addition of 300 nM DAMGO. The fluorescent signal observed after nigericin incubation may reflect the signal when the

16 membrane potential of the cells is driven to EK (Figure 5).

115

Figure 2: (A) Trace of fluorescent signal illustrating the reversal of DAMGO stimulated decrease in signal by MOR antagonist naloxone. 1 µM DAMGO/HBSS was added to AtT- 20MOR cells at 120 secs, followed by addition of 1 µM naloxone at 240 secs. The DAMGO stimulated decrease in fluorescent signal was completely reversed by naloxone. Naloxone had no effect when applied without previous addition of DAMGO. (B) Concentration response curve for morphine, both alone and with the addition of 3 nM, 10 nM and 100 nM naloxone. The addition of naloxone produced a parallel shift in the morphine concentration response curve with a pA2 of -8.5 ± 0.1, (2.9 ± 0.5 nM, n=3).

116 A 1000 RFU

800

600 1 mM Morphine/HBSS

400 1 mM Naloxone

200 Morphine + Naloxone HBSS + Naloxone

0 0 60 120 180 240 300 360 Time (secs)

B D Fluorescence (%)

30

20

0 nM Naloxone

10 3 nM Naloxone 10n nM Naloxone 100 nM Naloxone

0 -9 -8 -7 -6 -5 -4 [Morphine] log M

Figure 2

117 We assessed the capacity of the fluorescence assay to reliably detect opioid ligands of differing efficacy by treating AtT-20MOR cells with a range of structurally distinct opioid ligands. All agonists tested produced a concentration-dependent decrease in cellular fluorescence. High efficacy agonists such as fentanyl, DAMGO and β-endorphin produced a maximum decrease in fluorescence of 35-40%, while morphine, buprenorphine and pentazocine were shown to be partial agonists (Figure 6). A rank order of ligand efficacy was established (Table 1).

We assessed the suitability of this assay for HTS by calculating the Z’ factor, a measure of the assay robustness. An assay with a Z’ factor of between 0.5 - 1 is an excellent assay in terms of signal/noise ratio and data reproducibility13. The Z’ factor for this assay was calculated in multiple experiments, with Z’ values of 0.6, 0.7 and 0.7, indicating that the assay is suitable for HTS.

A hallmark of opioid signalling is agonist-dependent desensitisation, where continuous application of agonist results in a relatively rapid decline in receptor activation. The decrease in fluorescence produced by application of high concentrations of morphine faded in the continued presence of agonist, reaching a plateau after about 30 minutes (Figure

7A). In order to assess whether the decline in fluorescence reflected a change in MOR signalling, we incubated cells with a high concentration of morphine (1 µM) and then challenged them with subsequent addition of 10 µM morphine. The response to 10 µM morphine declined significantly over time. When fit with a one phase exponential association function, the peak response to 10 µM morphine declined with a τ of 490 s (95

% C.I. 413–603 s) to a maximum inhibition of 72 % (95 % C.I. 67–76 %, , Figure 7B). In order to assess whether the decline in morphine effectiveness reflected heterologous desensitisation, we repeated the experiments using 1 µM somatostatin as the challenge

118 drug. The response to somatostatin also declined during continuous morphine exposure (τ of 460 s, 95 % C.I. 331-748 s), the maximum inhibition of the somatostatin response was

30 % (95 % C.I. 27-34 %). Preincubation with naloxone (1 µM) did not affect the hyperpolarisation produced by somatostatin (10 nM; control 36 ± 1 %, in naloxone 35 ± 1

%).

Table 1: Potencies and efficacies of the range of structurally distinct opioid ligands tested using the membrane potential assay in AtT-20 cells. Ligands are ranked in order of efficacy. Ligands with Emax significantly different to that of DAMGO are marked with * (P < 0.05, extra sum of squares F test). Latency is the time taken for the peak signal to be reached after a maximal concentration of drug was added.

OPIOID pEC E (%) Hill Slope Latency (s) AGONIST 50 max

DAMGO 8.3 ± 0.1 38 ± 2 0.8 ± 0.1 29 ± 1 Endomorphin 2 8.4 ± 0.1 37 ± 2 1.0 ± 0.1 32 ± 2 Endomor phin 1 8.7 ± 0.1 36 ± 1 0.9 ± 0.1 27 ± 1 B-endorphin 7.0 ± 0.1 36 ± 3 1.2 ± 0.2 33 ± 4 Fentanyl 9.3 ± 0.1 35 ± 1 1.0 ± 0.1 27 ± 2 Met-Enkephalin * 8.5 ± 0.1 34 ± 1 0.8 ± 0.1 30 ± 3 Methadone * 7.6 ± 0.1 32 ± 1 1.2 ± 0.2 29 ± 3 Morphine * 7.7 ± 0.1 31 ± 1 1.0 ± 0.2 29 ± 2 Oxycodone * 6.7 ± 0.1 28 ± 1 1.2 ± 0.1 24 ± 3 Buprenorphine * 8.0 ± 0.1 21 ± 1 0.8 ± 0.3 12 ± 4 Pentazocine * 6.2 ± 0.2 5 ± 1 1.6 ± 0.3 18 ± 2

119 D Fluorescence (%) 40

30

20

10

0 DAMGO DAMGO + PTX Morphine Morphine + PTX Figure 3 Figure 3: The decrease in fluorescent signal observed with 30 nM DAMGO or 100 nM morphine was abolished following overnight incubation of AtT-20-MOR cells with 200ng/mL PTX.

D Fluorescence (%) 40

30

20

Low [K+] High [K+] 10

0 -10 -9 -8 -7 -6 [DAMGO] log M Figure 4 .

Figure 4: Decreasing extracellular potassium concentration ([K]Ex) increases the maximum change in fluorescent signal observed with DAMGO. Concentration- response curves were generated for DAMGO activation of GIRK channels in both

low [K]Ex (2.9 mM) and high [K]Ex (5.5 mM) conditions. Emax in low [K]Ex was 38 ±

2%, 21% higher than in high [K]Ex (Emax 31± 1%; P < 0.01).

120 RFU

600

500

400

300

200

1 mM Nigericin/HBS

100 300 nM DAMGO

0 0 300 600 900 1200 1500 1800 2100 Time (secs) Figure 5 Figure 5: Treatment of AtT-20-MOR cells with nigericin, a K selective antibiotic ionophore, caused a greater decrease in fluorescent signal than the maximally effective concentration of DAMGO (P < 0.05). 1 µM nigericin (heavy trace), or vehicle (light trace) was added to cells at 120 secs. Nigericin caused a decrease of 50 ± 3% in fluorescent signal. Membrane potential was allowed to reach equilibrium before the addition of 300 nM DAMGO at 1500 sec, which did not cause any further decrease in signal.

D Fluorescence (%) 40

DAMGO 30 Morphine Buprenorphine Pentazocine 20

10

0 -10 -9 -8 -7 -6 -5 -4 [opioid] log M Figure 6

Figure 6: Concentration response curves for DAMGO, morphine, buprenorphine and pentazocine activation of GIRK channels, illustrating the differences in agonist potency and efficacy. Morphine, buprenorphine and pentazocine all showed partial agonist activity with Emax values of 31 ± 1%, 21 ± 1% and 5 ± 1% respectively.

121

Figure 7: Desensitisation of MOR signalling in AtT-20 cells. Continuous application of morphine (1 µM) reduces the response to a subsequent addition of a high concentration of morphine or SRIF. Panel A shows example traces from an experiment where a high concentration of morphine (10 µM) or SRIF (1 µM) was added 30 minutes after the desensitizing concentration of 1 µM morphine (heavy trace) or 30 minutes after the addition of vehicle (light trace). Panel B shows the time course of the decline in response to 10 µM morphine or 1 µM SRIF in the continued presence of 1 µM morphine. Data are expressed as a percentage of the control response to drug added at the same time after the run commenced, and represent the mean ± s.e.m. of 3 - 8 determinations, each in duplicate or triplicate. The data was fit with a single exponential function to derive a t/2 for desensitisation (490s for morphine/morphine, 460s for morphine/somatostatin).

122 A

RFU RFU 800 800

600 600

400 1 mM Morhpine/HBSS 400 10 mM Morhpine 1 mM Morhpine/HBSS 1 mM SRIF 200 200

0 0 0 300 600 900 1200 1500 1800 2100 2400 0 300 600 900 1200 1500 1800 2100 2400

Time (secs) Time (secs)

B

Reduction of control response (%) 80

70

60

50

40

30

20

10

0 0 300 600 900 1200 1500 1800 2100 2400

Time after addition of morphine (1 mM)

Figure 7

123 DISCUSSION

In this study, we have developed a real time, no wash, fluorescence based assay for MOR- mediated hyperpolarisation of intact AtT-20 cells. We used a proprietary membrane potential sensitive dye from Molecular Devices to measure membrane hyperpolarisation following MOR mediated GIRK channel activation12. The fluorescent signal rapidly and reliably decreased after opioid application, and this decrease in signal could be completely reversed by the opioid naloxone applied for up to at least 2 hours after agonist, indicating the stability of the dye signal and capacity of the systems to report continued activation of the µ-receptors for prolonged periods, albeit in the face of receptor desensitisation. The ability to continuously measure the consequences of opioid receptor activation in cells for such prolonged periods of time has only previously been possible using high resistance electrode recordings from single neurons in brain slices7,17. Activation of endogenous sst receptors in AtT-20 cells15 produced a reduction in fluorescence similar that seen following MOR activation, and in both cases this reduction was less than that produced by application of the K-selective ionophore nigericin16.

The potency and efficacy of the opioid agonists correlate well with previous studies in

AtT-20 cells, and with studies in native neurons where activation of GIRK channels or

11,18,19,20 inhibition of ICa was used to determine agonist intrinsic activity . The data we obtained with the hyperpolarisation assay is similar to that we obtained when we measured

MOR inhibition of calcium currents in these cells – the rank order of potency in both studies is DAMGO > methadone = morphine > pentazocine11, and in both studies morphine and pentazocine have reduced efficacy when compared with DAMGO. The only notable inconsistencies are the apparently greater efficacy of endomorphin 1 and 2 compared with methadone in the present study. Many studies have reported the as being partial agonists18,21, while methadone has been reported to be an

124 agonist with a similar efficacy to DAMGO11,22,23. The most likely explanation for the discrepancy in our study is that methadone appeared to have a reduced maximal effect because of its propensity to block K channels, including the GIRK channels likely to contribute to the hyperpolarisation measured in the present study22,24. Conversely, the observation that the maximum effect of the endomorphins was similar to that of well recognized high efficacy agonists suggests the presence of some spare receptors in our system, as noted previously with similar cell lines11. It is also possible that the efficacy discrepancies reflect subtle bias in ligand signalling to one pathway over another in different tissues, and it is worth noting that endomorphins have recently been reported to show such bias in assays of β-arrestin recruitment25,26.

Continuous application of opioid agonists usually results in desensitisation of receptor signalling, a complex process that may involve receptor phosphorylation and sequestration.

The apparent stability of the membrane potential assay led us to explore whether it could be used to investigate µ-opioid receptor desensitisation. The hyperpolarisation produced by high concentrations of morphine or DAMGO appeared to wane over time, and when the cells were challenged with a concentration of agonist that should saturate the cell surface receptors there was a marked decrease in this response after only a few minutes exposure to agonist, consistent with analogous studies performed in locus coeruleus neurons27 or cell lines transfected with µ-opioid receptors and GIRK channels28. We were able to make use of the endogenous sst receptors in the AtT-20 cells to determine whether exposure to desensitizing concentrations of opioid agonist affected signalling through other receptors.

The time constant for the desensitisation of signalling produced by morphine was slower than that reported in our previous study of opioid signalling which utilized inhibition of ICa in AtT-20 cells, this may reflect a distinct recruitment of desensitisation processes by morphine in intact cells compared with cells dialyzed by patch-clamp recording11. A

125 major limitation of using the Flexstation for desensitisation assays is the inability to wash off drugs, which makes it impossible to do experiments where receptor function is repeatedly probed with concentrations of drugs that activate only a portion of receptors11,27.

Nevertheless, the potential of this assay to quickly test multiple agonists or putative modulators of receptor desensitisation in an intact system makes it an attractive option.

GIRK channel activation has most commonly been assessed using electrophysiological techniques, but assays potentially suitable for high throughput screening have been reported utilizing thallium flux29 or commercially available membrane potential dyes30,31.

Thallium is toxic and apparently unsuitable for all cell lines31 while other membrane potential sensitive dyes (Di-Bac, HLB 021-152) give results qualitatively similar to those

30,31 reported here . The EC50 value for the somatostatin-induced hyperpolarisation of our

AtT-20 cells was 2 nM, very similar to that previously reported using another dye (4 nM)31. However, it should be noted that under our standard conditions the proprietary dye gives a change in fluorescence of approximately 40% following GPCR activation, which compares with changes of approximately 10% using DiBAC4 in HL-1 or HLB 021-152 in

AtT-20 cells30,31. Removal of media and/or washing or the cells is also required during dye loading of commercial dyes used previously, this increases assay time and may promote cell detachment from the microplate wells31,32.

This assay has a number of advantages as a rapid screen for assessing ligand potency and efficacy at µ-opioid receptors. It is rapid, real-time, only requires the addition of a single reagent and is it performed in intact cells. When compared with assays of Gα subunit activation – either [35S]GTPγS binding assays or assays which directly measure agonist- stimulated GTPase activity33, the GIRK assay is far simpler, requires no handling of radioactivity and provides a real time measure of receptor activation rather than a single

126 point determination. However, the Gα subunit activation assays will provide a more sensitive discriminator of efficacy in cells with a low to moderate expression of receptors because the assay is constrained only by the amount of accessible Gα subunit in the cells, and this is usually not limiting33. By contrast, assays of AC regulation via measurement of cAMP accumulation have been adapted for use in whole cells, however these assays usually require incubation steps and either cell lysis followed by addition of several reagents or the use of cells transfected with which catalyse the production of a fluorescent substrate or with fluorescently labelled reporter proteins10. Assays of AC activity can be very sensitive and also detect cAMP levels over a large range of concentrations but in general when assaying the activity of Gi/Go-coupled receptors such as the µ-opioid receptor it is usually necessary to artificially elevate cAMP levels with forskolin in order to obtain an appropriate signal window for determining AC inhibition by the Gi/Go-derived Gα subunits. Nevertheless, assays of AC inhibition represent the most flexible and straightforward way of studying Gα (as opposed to Gβγ) signalling in a high throughput environment.

While both GTPγS binding and cAMP accumulation assays have been widely used for studying desensitisation of opioid receptor signalling, the usual requirement for control cells to be incubated with agonist for at least 10-15 minutes makes interpretation of the data problematic, as receptor desensitisation will be occurring during this time (e.g. Figure

7)7.

Other relatively straightforward strategies for real-time measurement of MOR-mediated

2+ signalling involve measuring [Ca]i concentration using Ca sensitive fluorescent dyes.

This approach provides relatively rapid real time response, and assays of [Ca]i concentration can be amenable to studying receptor desensitisation34,35. Interestingly, we

127 were unable to activate intracellular Ca release in AtT-20 cells via MOR or any of the Gαq receptors reportedly expressed in this cell line36 (data not shown).

This assay offers an alternative approach for measuring MOR activation and desensitisation by targeting a naturalistic Gβγ-mediated signalling pathway. Our results show the membrane potential assay to be a rapid, reliable and inexpensive method for identifying ligands that modulate GIRK activity, and may be scaled up to enable high- throughput screening for novel opioid drugs.

Acknowledgements

AK and MS were supported by MQRES scholarships, with top-ups from ASAM. PRB was an ASAM Summer Scholar. This work was supported by NHMRC Project Grant

APP1011979 to MJC and MC.

128 REFERENCES

1: Buntin-Mushock C, Phillip L, Moriyama K, Palmer PP: Age-dependent opioid escalation in chronic pain patients. Anesth Analg 2005; 100:1740-1745.

2: Matthes HWD, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I, Befort K,

Dierich A, Le Meur M, Dolle P, Tzavara E, Hanoune J, Roques BP, Kieffer BL: Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the

μ-opioid-receptor gene. Nature 1996; 383:819-823.

3: Alexander SPH, Mathie A, Peters JA: Guide to receptors and channels (GRAC), 5th edn. Brit J Pharmacol 2011; 164 (Suppl.1):S1-S324.

4: Law PY, Wong YH, Loh HH: Molecular mechanisms and regulation of opioid receptor signalling. Ann Rev Pharmacol Toxicol 2000; 40:389-430.

5: Rajagopal S, Rajagopal K, Lefkowitz RJ: Teaching old receptors new tricks: biasing seven-transmembrane receptors. Nat Rev Drug Disc 2010; 9:373-386.

6: Bailey CP, Llorente J, Gabra BH, Smith FL, Dewey WL, Kelly E, Henderson G: Role of protein kinase C and mu-opioid receptor (MOPr) desensitisation in tolerance to morphine in rat locus coeruleus neurons. Eur J Neurosci 2009; 29:307-318.

7: Connor M, Osborne PB, Christie MJ: mu-Opioid receptor desensitisation: Is morphine different? Brit J Pharmacol 2004; 143:685-696.

8: Pradhan AA, Walwyn W, Nozaki C, Filliol D, Erbs E, Matifas , Evans C, Kieffer BL:

Ligand-directed trafficking of the delta-opioid receptor in vivo: two paths toward analgesic tolerance. J Neurosci 2010; 30:16459-16468.

129 9: Coward P, Chan SDH, Wada HG, Humphries GM, Conklin BR: Chimeric G-proteins allow a high-throughput signalling assay of Gi-coupled receptors. Anal Biochem 1999;

270:242-248.

10: Hill SJ, Williams C, May LT: Insights into GPCR pharmacology from the measurement of changes in intracellular cyclic AMP; advantages and pitfalls of differing methodologies. Brit J Pharmacol 2010; 161:1266-1275.

11: Borgland SL, Connor M, Osborne PB, Furness JB, Christie MJ: Opioid agonists have different efficacy profiles for G protein activation, rapid desensitisation, and endocytosis of mu-opioid receptors. J Biol Chem 2003; 278:18776-18784.

12: Celver J, Xu M, Jin W, Lowe J, Chavkin C: Distinct domains of the μ control uncoupling and internalisation. Mol Pharmacol 2004; 65:528-537.

13: Zhang J-H, Chung TDY, Oldenburg KR: A Simple Statistical Parameter for Use in

Evaluation and Validation of High Throughput Screening Assays. J Biomol Screen 1999;

4:67-73.

14: Kaufman DL, Keith Jr DL, Anton B, Tian J, Magendzo K, Newman D, Tran TH, Lee

DS, Wen C, Xia Y-R, Lusis AJ, Evans CJ: Characterisation of the murine µ-opioid receptor gene. J Biol Chem 1995; 270: 15877-15883.

15: Cervia D, Nunn C, Fehlmann D, Langenegger D, Schuepbach E, Hoyer D:

Pharmacological characterisation of native somatostatin receptors in AtT-20 mouse tumour corticotrophs. Brit J Pharmacol 2003; 139:109-121.

16: Doebler JA: Comparative effects of carboxylic ionophores on membrane potential and resistance of NG108-15 cells. Toxicol in Vitro 2000; 14:235-243.

130 17: Pepper CM, Henderson G: Opiates and opioid peptides hyperpolarize locus coeruleus neurons in vitro. Science 1980; 209:394-395.

18: Connor M, Schuller A, Pintar JE, Christie MJ: μ calcium channel current in periaqueductal grey neurons from C57B16/J mice and mutant mice lacking MOR-1 Brit J Pharmacol 1999; 126:1553-1558.

19: Christie MJ, Williams JT, North RA: Cellular mechanisms of opioid tolerance: studies in single brain neurons. Mol Pharmacol 1987; 32:633-638.

20: Ingram S, Wilding TJ, McCleskey EW, Williams JT: Efficacy and kinetics of opioid action on acutely dissociated neurons. Mol Pharmacol 1997; 52:136-143.

21: Alt A, Mansour A, Akil H, Medzihradsky F, Traynor JR, Woods JH: Stimulation of

Guanosine-5'-O-(3-[35S]Thio)Triphosphate binding by endogenous opioids acting at a cloned mu-receptor. J Pharmacol Exp Ther 1998; 286:282-288.

22: Rodriguez-Martin I, Braksator E, Bailey CP, Goodchild S, Marrion NV, Kelly E,

Henderson G: Methadone: does it really have low efficacy at μ-opioid receptors?

Neuroreport 2008; 19:589-593

23: Selley DE, Liu Q, Childers SR: Signal transduction correlates of mu opioid agonist intrinsic efficacy: receptor-stimulated [35S]GTPγS binding in mMOR-CHO cells and rat thalamus. J Pharmacol Exp Ther 1998; 285:496-505.

24: Matsui A, Williams JT: Activation of µ-opioid receptors and block of KIR3 potassium channels and NMDA receptor conductance by l- and d-methadone in rat locus coeruleus.

Brit J Pharmacol; 161:1403-1413.

131 25: McPherson J, Rivero G, Baptist M, Llorente J, Al-Sabah S, Krasel C, Dewey WL,

Bailey CP, Rosethorne EM, Charlton SJ, Henderson G, Kelly E: µ-Opioid receptors: correlation of agonist efficacy for signalling with ability to activate internalisation. Mol

Pharm 2010; 78:756-766.

26: Rivero G, Llorente J, McPherson J, Cooke A, Mundell SJ, McArdle CA, Rosethorne

EM, Charlton SJ, Krasel C, Bailey CP, Henderson G, Kelly E: Endomorphin-2: a biased agonist at the µ-Opioid receptor. Mol Pharm 2012; 82, 178-188.

27: Harris G, Williams J: Transient homologous mu-opioid receptor desensitisation in rat locus coeruleus neurons. J Neurosci1991; 11:2574-2581.

28: Johnson EA, Oldfield S, Braksator E, Gonzalez-Cuello A, Couch D, Hall KJ, Mundell

SJ, Bailey CP, Kelly E, Henderson G: Agonist-selective mechanisms of mu-opioid receptor desensitisation in human embryonic kidney 293 cells. Mol Pharmacol 2006;

70:676-685.

29: Niswender CM, Johnson KA, Luo Q, Ayala JE, Kim C, Conn PJ, Weaver CD: A novel assay of Gi/o-linked G protein-coupled receptor coupling to potassium channels provides new insights into the pharmacology of the Group III metabotropic glutamate receptors.

Mol Pharmacol 2008; 73:1213-1224.

30: Walsh KB: A real-time screening assay for GIRK1/4 channel blockers. J Biomol

Screen 2010; 15:1229-1237.

31: Walsh KB: Targeting GIRK channels for the development of new therapeutic agents.

Front Pharmacol 2011; 2:1-8.

132 32: Wolff C, Fuks B, Chatelain P: Comparative Study of Membrane Potential-Sensitive

Fluorescent Probes and their Use in Ion Channel Screening Assays. J Biomol Screen 2003;

8:533-543.

33: Harrison C, Traynor JR: The [35S]GTPγS binding assay: approaches and applications in pharmacology. Life Sci 2003; 74:489-508.

34: Connor MA, Keir MJ, Henderson G: δ-opioid receptor mobilisation of intracellular calcium in SH-SY5Y Cells: lack of Evidence for δ-receptor subtypes. Neuropharmacology

1997; 36:125-133.

35: Chu J, Zheng H, Zhang YH, Loh HH, Law PY: Agonist-dependent mu-opioid receptor signalling can lead to heterologous desensitisation. Cell Signall 2010; 22:684-696.

36: Atwood BK, Lopez J, Wager-Miller J, Mackie KAS: Expression of G protein-coupled receptors and related proteins in HEK293, AtT-20, BV8 and N18 cell lines as revealed by microarray analysis. BMC Genomics 2011; 12:14.

133

134 RESULTS

The following chapters describe the consequences of MOPr SNPs on receptor signalling via a range of clinically important and/or structurally distinct opioid ligands. Chapter 5,

“Buprenorphine signalling is compromised at the N40D polymorphism of the human µ- opioid receptor in vitro", describes how the common N40D variant significantly reduces maximum buprenorphine inhibition of AC and phosphorylation of ERK1/2, and reduces buprenorphine potency for GIRK activation. Chapter 6, “The A6V polymorphism of the human µ-opioid receptor negatively impacts signalling of morphine and endogenous opioids in vitro” reports the aberrant signalling of a range of opioid ligands at the relatively common A6V MOPr variant. Chapter 7, “Mu-opioid receptor polymorphisms differentially affect receptor signalling via adenylyl cyclase inhibition and ERK phosphorylation”, describes the effects of the L85I, R260H and R265H variants on the ability of MOPr to couple to pathways of AC inhibition and ERK1/2 phosphorylation.

135

136 Chapter 5

Buprenorphine signalling is compromised at the N40D polymorphism of the human µ-opioid receptor in vitro

Alisa Knapman, Marina Santiago and Mark Connor

This paper was submitted for peer review to the British Journal of Pharmacology on

23/4/14. Marina Santiago constructed AtT20-MOPr cells and performed GIRK activation assays. Mark Connor assisted with experimental design, data analysis and manuscript preparation. All other work is my own.

137 SUMMARY

Background and Purpose

There is significant variation in individual response to opioid drugs, which may result in inappropriate opioid therapy. Polymorphisms of the µ-opioid receptor (MOPr) may contribute to individual variation in opioid response by affecting receptor function, and the effect may be ligand-specific. We sought to determine functional differences in MOPr signalling at several signalling pathways using a range of structurally distinct opioid ligands in cells expressing wild-type MOPr and the commonly occurring MOPr variant,

N40D.

Experimental Approach

MOPr-WT and MOPr-N40D were stably expressed in CHO cells and in AtT-20 cells.

Assays of adenylyl cyclase inhibition and ERK1/2 phosphorylation were performed on

CHO cells, and assays of K activation were performed on AtT-20 cells. Signalling profiles for each ligand were compared between variants.

Key Results

Buprenorphine efficacy was reduced by over 50% at MOPr-N40D for adenylyl cyclase inhibition and ERK1/2 phosphorylation. Buprenorphine potency was reduced threefold at

MOPr-N40D for K channel activation. Pentazocine efficacy was reduced by 50% for

GIRK activation at MOPr-N40D. No other differences were observed for any other ligands tested.

Conclusions and Implications

The N40D variant is present in 10 - 50% of the population. Buprenorphine is a commonly prescribed opioid analgesic, and many individuals do not respond to buprenorphine therapy. This study demonstrates that buprenorphine signalling to several effectors via the

N40D variant of MOPr is impaired, and this may have important consequences in a clinical setting for individuals carrying the N40D allele.

138 Abbreviations

AC adenylyl cyclase

BSA bovine serum albumin

CHO Chinese Hamster Ovary

DAMGO ([D-Ala2, N-MePhe4, Gly-ol]-enkephalin)

DMEM Dulbecco’s Modified Eagle Medium

ERK extracellular signal regulated kinase

FSK forskolin

GIRK G protein-gated inwardly rectifying K channel

GPCR G protein couple receptors

HBSS Hanks balanced salt solution

MOPr µ-opioid receptor

OPRM1 opioid receptor mu 1

PTX pertussis toxin

139 INTRODUCTION

Opioids are powerful analgesics used for the clinical management of moderate to severe pain. Opioid use is associated with adverse effects such as respiratory depression, nausea, constipation and sedation, and tolerance and dependence can develop with continued opioid use. There is significant variation between individuals in response to opioid drugs, both in the analgesic and adverse effects (Lotsch et al., 2005). This variation can lead to restricted dosing of opioid analgesics due to the unpredictability of serious adverse events such as respiratory depression, resulting in inadequate pain relief for many individuals

(Skorpen & Laugsand, 2008). The inter-individual variability observed in response to opioids is likely to be caused, at least in part, by genetic differences in proteins responsible for drug absorption, distribution and metabolism, as well as differences in drug/receptor interactions and receptor signalling (Somogyi et al., 2007). A clearer understanding of the genetic factors contributing to individual response to opioids could result in the ability to better predict the outcomes of opioid administration in individuals, leading to more rational choice of drug and dose, thus potentially limiting adverse effects and the development of tolerance and dependence.

The µ-opioid receptor (MOPr, Cox et al., 2014) is a G-protein coupled receptor (GPCR) that is the main site of action for clinically important opioids. MOPr mediates the analgesic and almost all of the adverse effects of these opioid drugs, and it is likely that genetic variation of OPRM1, the gene coding for MOPr, could contribute to individual variation in the response to opioid analgesics. A number of non-synonymous allelic variants of

OPRM1 have been identified within the population, each resulting in an alternative receptor isoform (Lotsch et al., 2005). The N40D variant is the most common MOPr variant, occurring at an allelic frequency of 10-50% in various populations (Mura et al.,

2013). This variant arises from an A > G substitution at nucleotide 118, resulting in an

140 asparagine to aspartic acid amino acid exchange in the N-terminal domain of MOPr, and removing a putative glycosylation site (Singh et al., 1997).

Many studies have investigated the association between carriers of MOPr-N40D and various clinical outcomes, such as the degree of pain relief from opioid analgesics, and the susceptibility to substance abuse. A number of these studies suggest that carriers of the

D40 allele require higher doses of post-operative opioid analgesics, however other studies have shown an increased sensitivity to opioids and a reduced perception of pain

(Diatchenko et al., 2011; Mura et al., 2013). There are also reports of an increased susceptibility to alcohol and heroin abuse in D40 carriers, but an improved response to naltrexone treatment of alcoholism (Mague & Blendy, 2010). Despite the volume of research into the effect of the N40D variant on disease outcomes, many of the reports are contradictory and there is no clear consensus as to the effect of the N40D variant on the outcomes of opioid administration or disease susceptibility (Walter & Lotsch, 2009).

Fewer studies have investigated the consequences of the N40D variant on MOPr signalling and function in vitro, and the results of these studies are also conflicting. One study reported a threefold increase in -endorphin affinity for MOPr heterologously expressed in

AV-12 cells, and a similar increase in -endorphin potency for G protein-gated inwardly rectifying K channel (GIRK) activation in Xenopus oocytes (Bond et al., 1998).

Subsequent studies have found no differences in -endorphin potency between N40D- and

WT-MOPr (Befort et al., 2001; Beyer et al., 2004; Kroslak et al., 2007). Other studies have reported both increased and decreased DAMGO and morphine potency, differences in regulatory processes with chronic opioid treatment, and brain region-dependent differences in expression and signalling, however there is little consistency in these reports

141 (reviewed in Knapman & Connor, 2014). As MOPr interacts with many effector and regulatory proteins, changing assay parameters and cellular background may affect MOPr signalling in different ways. Furthermore, most studies have used only a single or a small subset of opioid ligands, and may have failed to capture ligand-specific differences in

N40D signalling.

Like all G protein couple receptors (GPCR), MOPr has many active conformations, and in the absence of ligand constantly oscillates through a range of possible active and inactive states (Pinyero et al., 2007; Kenakin & Miller, 2010; Manglik et al., 2012). Ligand-biased signalling or functional selectivity has been well characterised, where structurally distinct ligands stabilise the receptor in a restricted subset of conformations, preferentially activating certain effectors. The corollary of this phenomenon is that changes in amino acid residues arising from genetic polymorphisms may also affect conformation of the receptor (Abrol et al., 2013; Cox, 2013). Thus, the N40D variant may affect MOPr signalling globally or in a ligand-dependent manner by affecting the ability of a ligand to bind to the receptor, altering the conformation of the ligand-receptor complex and/or affecting the ability of this complex to couple to G-proteins and associated signalling or regulatory pathways. Clinically used opioids are chemically diverse, and are likely to have subtle differences in their interaction with structural features of the MOPr, potentially leading to distinct effects of the N40D variant on different drugs. In this study, the ability of human MOPr-WT and MOPr-N40D to couple to several distinct signalling pathways was investigated in Chinese hamster ovary (CHO) cells and mouse AtT-20 cells using 11 clinically important and/or structurally distinct opioid ligands. We found that the N40D variant had a negative impact on the signalling of the commonly prescribed opioid buprenorphine in all of our assays. In addition, the efficacy of pentazocine was

142 significantly reduced for GIRK activation in AtT-20 cells expressing N40D. No differences in signalling via other opioids, including -endorphin, were observed.

METHODS

MOR transfection and cell culture

CHO-FRT-TRex cells were stably transfected with a pcDNA5 construct encoding the haemagglutinin-tagged human -opioid receptor cDNA together with the pOG44 (Flp recombinase plasmid) using the transfectant Fugene (Promega), as described in previously

(Knapman et al., 2014). The HA-tagged human wild-type -opioid receptor (MOPr-WT) and the -opioid receptor containing the D40 variant (MOPr-N40D) were synthesised by

Genscript (Piscataway, New Jersey, USA). Cells expressing MOPr-WT or MOPr-N40D were selected using hygromycin B (500 µg/mL). Cells were cultured in Dulbecco’s

Modified Eagle Medium (DMEM) containing 10% FBS, 100U penicillin/streptomycin and

500 g/ml hygromycin B up to passage 5. Hygromycin concentration was reduced to 200

g/ml beyond passage 5.

AtT-20 FlpIn cells were constructed by transfecting Flp-recombinase target site

(FRT)/LacZeo2 (Life Technologies) using Fugene. Successfully transfected cells were selected with 100 µg/ml zeocin, and stably transfected with MOPr-WT or MOPr-N40D as described for CHO cells. Cells expressing MOPr were selected using 100 µg/ml hygromycin B. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% FBS, 100U penicillin/streptomycin and 100 g/ml hygromycin B.

CHO-MOPr cells and AtT-20-MOPr cells were passaged at 80% confluency as required.

Assays were carried out on cells up to 30 passages. Cells for assays were grown in 75 cm2 flasks and used at greater than 90% confluence.

143 MOPr receptor expression

Surface expression of MOPr was determined on intact CHO-MOPr cells by incubation with 0.125 – 16 nM [3H]-DAMGO (D-Ala2, N-MePhe4, Gly-ol]-enkephalin; PerkinElmer,

Waltham, MA) at 4 °C in 50 mM Tris-Cl (pH 7.4) for 2 h. Briefly, 24 hours before the assay, cells were detached from flasks with trypsin/EDTA (Sigma) and resuspended in

DMEM containing 10% FBS, 100U penicillin/streptomycin, plus 2 µg/ml tetracycline to induce MOPr expression. Cells were seeded at a density of 1 × 105 cells per well in 24- well plates pre-coated with polylysine and grown overnight at 37° C. On the day of the assay, cells were washed twice gently with 50 mM Tris-Cl (pH 7.4) and incubated for 2 h on ice with 0.125 – 16 nM [3H]-DAMGO. Non-specific binding was determined in the presence of unlabeled DAMGO (10 µM). Non-specific binding was 15 ± 1% of total binding for CHO-MOPr-WT, and 11 ± 1% in CHO-MOPr-N40D. At the end of the incubation, plated cells were washed three times with 50 mM Tris-Cl (pH 7.4) at 4 °C.

Cells in each well were then digested for 1 h at room temperature with 100 µl of 1N

NaOH. 100 µl 1N HCl was added to each well and collected into scintillation vials and bound ligand determined using a liquid scintillation counter (Packard Tricarb, Perkin

Elmer,Waltham MA, USA). Receptor density (Bmax) and affinity (Kd) were calculated using a one-site binding curve fitted using GraphPad Prism (GraphPad Software, La Jolla,

CA). Protein concentration was determined with a BCA Protein Assay Kit (Pierce) according to the manufacturer’s instructions. All experiments were performed three times in triplicate.

144 Membrane Potential Assay of Adenylyl Cyclase Inhibition

Opioid inhibition of adenylyl cyclase (AC) was measured using an assay of membrane potential (Knapman et al., 2014a). The day before the assay, CHO-MOPr cells were detached from the flask with trypsin/EDTA (Sigma) and resuspended in 10 ml of

Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin and 15 mM glucose. MOPr receptor expression was induced with 2 g/ml tetracycline 24 hours prior to the assay. The cells were plated in a volume of 90 µl in black walled, clear-

o bottomed 96-well microplates (Corning), and incubated overnight at 37 C in ambient CO2.

Membrane potential was measured using a FLIPR Membrane Potential Assay kit (blue) from Molecular Devices. The dye was reconstituted with assay buffer (Hanks balanced salt solution, HBSS) containing (in mM), NaCl 145, HEPES 22, Na2HPO4 0.338, NaHCO3

4.17, KH2PO4 0.441, MgSO4 0.407, MgCl2 0.493, CaCl2 1.26, glucose 5.56 (pH 7.4, osmolarity 315 ± 5). Prior to the assay, cells were loaded with 90 l/well of the dye solution without removal of the L-15, giving an initial assay volume of 180 l/well. Plates were then incubated at 37°C in ambient CO2 for 60 minutes. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at 565 nm. Baseline readings were taken every 2 seconds for at least 2 minutes, at which time forskolin (FSK, an activator of AC) and either opioid or vehicle was added in a volume of 20 l. The background fluorescence of cells without dye or dye without cells was negligible. Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle addition. The final concentration of DMSO was not more than 0.1%, and this concentration did not produce a signal in the assay.

145 Membrane Potential Assay of GIRK Channel Activation

Opioid activation of endogenous GIRK channels in AtT-20 cells was measured using a membrane potential sensitive dye, as previously described (Knapman et al., 2013). AtT-

20-MOPr cells were detached from flasks and plated using the same procedure as for the

CHO-MOPr cells in the AC inhibition assay, with no addition of tetracycline. Blue membrane potential dye was reconstituted and loaded onto cells, and the assay was performed in the same way as the AC inhibition assay. Baseline readings were taken every

2 seconds for at least 2 minutes, at which time opioid or vehicle was added in a volume of

20 l. Measurements were taken at the peak decrease in signal from baseline, approximately 10-20 sec after drug addition. The background fluorescence of cells without dye or dye without cells was negligible. The final concentration of DMSO was not more than 0.1%, and this concentration did not produce a signal in the assay.

ELISA of ERK1/2 phosphorylation

Opioid-induced phosphorylation of extracellular signal regulated kinase 1/2 (ERK1/2) was measured by ELISA. The day before the assay, CHO-MOPr cells were plated and receptor expression was induced as for the membrane potential assay. Cells were plated in 96-well clear microplates (Falcon). On the day of the assay, cells were serum-starved for 1 hr in 40

µl serum-free L-15 supplemented with 5% bovine serum albumin (BSA). All cell treatments were in a volume of 40 µl unless stated otherwise. Serum-starved cells were treated with drug or vehicle diluted in serum-free L-15. Preliminary experiments indicated that drug treatment for 5 min was optimal to produce robust ERK1/2 phosphorylation without inducing desensitisation, thus all drug treatments were for 5 min unless otherwise indicated. After drug application, the reaction was stopped by inverting the plates to remove the drug solution, placing the plates on ice, and immediately fixing the cells with

4% paraformaldehyde for 15 min at RT. Cells were washed 3 times with 300 µl PBS, then 146 permeabilized with 0.1% Triton-X in PBS for 30 min at RT. Triton-X was removed, and cells were incubated for 2 hr at RT with blocking solution consisting of 5% BSA in PBS with 0.01% Tween-20 (PBS-T). Blocking solution was removed, then cells were incubated overnight at 4°C with a 1:500 dilution of rabbit anti-phospho-p44/42 MAPK

(Thr202/Tyr204) antibody in PBS-T with 1% BSA. Cells were washed 3 times with 300 µl

PBS-T, and incubated with 1:5000 anti-rabbit IgG HRP-linked antibody in PBS-T with 1%

BSA for 2 hr. Cells were washed 4 times with 300 µl PBS-T, and incubated with 3,3’,5,5’- tetramethylbenzidine (Sigma) at RT in the dark for 45 min. The reaction was stopped with

1M HCl. Absorbance was read at 450 nm using a BMG Pherastar FS microplate reader.

Cells were then stained with 0.5 µg/mL DAPI for 10 min at RT, and washed 3 times with

300 µl PBS-T. Fluorescence was read in the Pherastar microplate reader with cells excited at a wavelength of 358 nm and emission measured at 461 nm. Absorbance readings were normalised to DAPI staining to account for differences in cell density between wells.

Readings were then normalised to the response of cells treated with 100 nM PMA for 10 min.

Drugs and Chemicals

Unless otherwise noted, tissue culture reagents and buffer salts were from Invitrogen or

Sigma. DAMGO, endomorphin-1, endomorphin-2 and met-enkephalin were purchased from Auspep (Tullamarine, Australia). Morphine and pentazocine were a kind gift from the Department of Pharmacology, University of Sydney. Buprenorphine and oxycodone were from the National Measurement Institute (Lindfield, Australia). -endorphin was from Genscript (Piscataway, New Jersey, USA). Fentanyl (Andrews Laboratories) was a gift from the Department of Pharmacology, Sydney University. Forskolin and naloxone were from Ascent Pharmaceuticals (Bristol, UK). 1,9-dideoxyforskolin and tetraethylammonium (TEA) were from Sigma Aldrich (Castle Hill, Australia). Pertussis 147 toxin (PTX) was from Tocris Bioscience (Bristol, UK). Phospho-ERK1/2 antibody

(Catalog #9101) and anti-rabbit IgG HRP-lined antibody (Catalog #7074) were from Cell

Signalling Technologies, Danvers, Massachusetts, USA.

Data

Unless otherwise noted, data is expressed as mean ± s.e.m. of at least 5 determinations made in duplicate or triplicate. Concentration response curves were fit with a 4 parameter logistic equation using Graphpad Prism (Graphpad). Emax and pEC50 values were derived from individual experiments and compared using unpaired Student’s T-test. P < 0.05 was considered significant. Full agonist activity was determined by comparing Emax values derived from individual experiments using one-way ANOVA, corrected for multiple comparisons. Pooled data from replicate experiments is shown in the Figures for ease of reference. All channel and receptor nomenclature is consistent with the British Journal of

Pharmacology/IUPHAR Concise Guide to Pharmacology (Alexander et al., 2013).

RESULTS

MOPr Expression in CHO-K1 cells

CHO-K1 cells were stably transfected with either MOPr-WT or the MOPr-N40D variant, with receptor expression controlled by a tetracycline-sensitive repressor. After induction of

MOPr expression with tetracycline, specific binding of [3H]DAMGO was similar for both variants, indicating similar levels of surface receptor expression (see Figure 1). The Bmax for CHO-MOPr-WT cells was 280 ± 20 fmol/mg total protein and 356 ± 18 fmol/mg for

CHO-MOPr-N40D (P > 0.05). The affinity for [3H]DAMGO was also similar between for

WT-MOPr and N40D-MOPr expressing cells, with KD of 0.75 ± 0.10 and 0.65 ± 0.2, respectively (P > 0.05).

148

Specific Binding (DPM) 2500

2000

1500 WT 1000 N40D

500

0 0510 15 20 [(3H)DAMGO] nM

Figure 1: Saturation binding curve of [3H]DAMGO in intact CHO-MOPr-WT and CHO-MOPr-N40D cells, 24 hr after induction of receptor expression with tetracycline. Radioligand binding was carried out as described in the Methods. No significant difference in Bmax or Kd was observed between cells expressing MOPr-WT or MOPr-N40D (P > 0.05). Each point represents the mean ± s.e.m. of triplicate determinations from a single experiment. The assay was repeated 3 times.

149 Adenylyl Cyclase Inhibition

In CHO-MOPr cells loaded with membrane potential dye, addition of the adenylyl cyclase activator forskolin (FSK, 300 nM) produced a rapid decrease in fluorescence, consistent with hyperpolarisation of the cells (Figure 2; Knapman et al., 2014). The hyperpolarisation stabilised 5 min after the addition of FSK, at which point measurements were taken. FSK (300 nM) produced a similar hyperpolarisation in CHO-MOPr-WT (42 ±

1 % decrease in fluorescence) and CHO-MOPr-N40D (44 ± 2 % decrease) cells. (P >

0.05). Simultaneous addition of the prototypical MOPr selective peptide agonist DAMGO with FSK resulted in a concentration-dependent inhibition of the FSK-stimulated hyperpolarisation in both cell lines (Figure 2). DAMGO maximally inhibited the FSK response by 60 ± 2 %, with pEC50 of 7.7 ± 0.1 in CHO-MOPr-WT cells (Table 1).

DAMGO inhibited FSK similarly in cells expressing N40D, Emax was 67 ± 6 % and pEC50

7.7 ± 0.1 (P > 0.05). We have reported previously that the effects of DAMGO in this assay were sensitive to naloxone and strongly inhibited by overnight pretreatment with pertussis toxin (Knapman et al., 2014). Application of opioids alone did not affect the membrane potential of CHO-MOPr cells with the exception of 10 µM methadone and 30 µM pentazocine, which both caused a small, transient increases in fluorescence (< 10%). These effects were not naloxone sensitive and were not observed at lower concentrations of drug.

They are consistent with previously reported inhibition of K channels by these drugs

(Matsui and Williams, 2010; Nguyen et al., 1998).

Because of the uncertainty surrounding the effects of N40D polymorphism on opioid actions, we measured the potency and efficacy of a range of clinically important and/or structurally distinct opioid ligands in CHO cells expressing WT-MOPr or the N40D variant to determine whether the N40D amino acid substitution could affect MOPr- signalling in a ligand-selective manner. The endogenous opioid -endorphin has

150 previously been shown to have a three-fold increase in potency at GIRK channel activation in Xenopus oocytes expressing N40D-MOPr (Bond et al., 1998). In the present assay, there was no difference in -endorphin potency or efficacy between CHO-MOPr-WT cells and

CHO-MOPr-N40D cells (Figure 3, Table 1). AC inhibition by the putative endogenous opioid ligands endomorphin-1 and endomorphin-2, as well as met-enkephalin were also similar in cells expressing MOPr-WT or MOPr-N40D receptor (Figure 3, Table 1).

The efficacy and potency of the prototypical opioid alkaloid morphine was not different in cells expressing the N40D variant (see Figure 4, Table 1). Interestingly, the semisynthetic morphine derivative buprenorphine had a significantly lower efficacy for AC inhibition in

CHO cells expressing MOPr-N40D when compared with MOPr-WT. The buprenorphine

Emax was 35 ± 6% in CHO-MOPr-WT cells, and 16 ± 4% in CHO-MOPr-N40D cells, a reduction of over 50% (P < 0.05). Buprenorphine potency was similar at the N40D variant

(see Figure 4, Table 1). The low efficacy of buprenorphine at MOPr could have accentuated a general difference in transduction efficiency between MOPr-WT and the

MOPr-N40D. However, the low efficacy MOPr agonist pentazocine inhibited AC to a similar degree in cells expressing MOPr-WT or MOPr-N40D (Figure 4, Table 1), indicating that the low efficacy of buprenorphine per se was not responsible for reduced signalling at the N40D variant. No difference in AC signalling was observed between cells expressing MOPr-WT or the MOPr-N40D variant when challenged with fentanyl, methadone or oxycodone (Figure 5, Table 1).

151 Figure 2: DAMGO inhibits adenylyl cyclase and activates ERK1/2 in CHO cells expressing MOPr-WT or MOPr-N40D. Adenylyl cyclase inhibition and levels of ERK1/2 phosphorylation were determined as described in the Methods. A) Traces showing changes in fluorescent signal following application of 300 nM forskolin + vehicle (Hank’s Balanced Salt Solution, HBSS) to MOPr-WT, and 300 nM forskolin + 1 µM DAMGO to MOPr-WT and MOPr-N40D cells. Drugs were added at 120 sec. Changes in RFU are normalized to predrug values. B) DAMGO inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). C) DAMGO stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

152 A

RFU (% baseline)

100

80

60

40 300 nM FSK + HBSS 20 300 nM FSK + 1 mMDAMGO (WT) 300 nM FSK + 1 mMDAMGO (N40D) 0 0120 240 360 480 600 720 Time (s) B Inhibition of FSK response (%) 100 WT 80 N40D 60

40

20

0 -9 -8 -7 -6 [DAMGO] log M C pERK1/2 (% Control) 100

80 WT N40D 60

40

20

0 -10 -9 -8 -7 -6 [DAMGO] log M

153 Figure 3: Endogenous opioids inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr-N40D. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) β-endorphin, endomorphins 1 & 2, and met-enkephalin inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). B) -endorphin, endomorphins 1 & 2, and met-enkephalin stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

154 A B Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT N40D N40D 60 60

40 40

20 20

0 0 -8 -7 -6 -5 -10 -9 -8 -7 -6 [b-Endorphin] log M [b-endorphin] log M Inhibition of FSK response pERK1/2 (% Control) 120 100 (%) 100 WT 80 WT N40D 80 N40D 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-1] log M [Endomorphin-1] log M pERK1/2 (% Control) Inhibition of FSK response 120 100 (%)

100 WT 80 WT N40D 80 N40D 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-2] log M [Endomorphin-2] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 120

100 80 WT WT N40D 80 N40D 60 60 40 40 20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Met-Enkephalin] log M [Met-Enkephalin] log M

155

Figure 4: Buprenorphine inhibits adenylyl cyclase and activates ERK1/2 less effectively in CHO cells expressing MOPr-N40D. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) Buprenorphine

Emax for inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation was decreased from 35 ± 6 % in CHO-MOPr-WT to 16 ± 4% in CHO-MOPr-N40D (P < 0.05). Morphine and pentazocine inhibited adenylyl cyclase hyperpolarisation to a similar degree and with similar potency in CHO cells expressing MOPr-WT and MOPr-N40D (P > 0.05).

B) Buprenorphine Emax for stimulation of ERK1/2 phosphorylation was decreased from 35 ± 7 % in CHO-MOPr-WT to 14 ± 6% in CHO-MOPr-N40D (P < 0.05). Morphine and pentazocine stimulated ERK1/2 phosphorylation to a similar degree and with similar potency in CHO cells expressing MOPr-WT and MOPr-N40D (P > 0.05). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

156 A B

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT N40D N40D 60 60

40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Morphine] log M [Morphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT N40D N40D 60 60

40 40

20 20

0 0 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 [Buprenorphine] log M [Buprenorphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

WT 80 WT 80 N40D N40D 60 60

40 40

20 20

0 0 -7 -6 -5 -4 -8 -7 -6 -5 -4 [Pentazocine] Log M [Pentazocine] Log M

157 Figure 5: Fentanyl, oxycodone and methadone inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr-N40D. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) Fentanyl, oxycodone and methadone inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). B) Fentanyl, oxycodone and methadone stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-N40D to a similar degree and with a similar potency (P > 0.05). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

158 A B

Inhibition of FSK response pERK1/2 (% Control) 120 (%) 120

100 100 WT WT 80 N40D 80 N40D

60 60

40 40

20 20

0 0 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Fentanyl] log M [Fentanyl] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT N40D N40D 60 60

40 40

20 20

0 0 -8 -7 -6 -5 -4 -8 -7 -6 -5 -4 [Oxycodone] log M [Oxycodone] log M

Inhibition of FSK response pERK1/2 (% Control) 100 100 (%)

80 WT 80 WT N40D N40D 60 60

40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Methadone] log M [Methadone] log M

159 Table 1: Summary of opioid efficacy and potency in assays of AC inhibition in CHO cells expressing MOPr-WT and MOPr-N40D.

AC inhibition Emax (%) pEC50 Opioid WT N40D WT N40D -Endorphin 70 ± 4 69 ± 6 6.7 ± 0.1 6.7 ± 0.1 Methadone 66 ± 2 73 ± 5 7.0 ± 0.1 7.2 ± 0.6 Met-enkephalin 66 ± 4 73 ± 4 7.8 ± 0.1 7.7 ± 0.3 Morphine 66 ± 3 70 ± 4 7.0 ± 0.1 7.0 ± 0.1 Fentanyl 65 ± 2 74 ± 4 8.1 ± 0.2 8.2 ± 0.1 Oxycodone 65 ± 5 77 ± 6 5.8 ± 0.3 6.0 ± 0.1 Endomorphin-1 62 ± 3 61 ± 4 8.3 ± 0.1 8.3 ± 0.1 Endomorphin-2 60 ± 7 64 ± 6 8.1 ± 0.1 8.3 ± 0.1 DAMGO 60 ± 2 67 ± 6 7.7 ± 0.1 7.7 ± 0.2 Pentazocine 50 ± 6 47 ± 6 5.2 ± 0.1 4.8 ± 0.6 35 ± 6 16 ± 4* 8.5 ± 0.2 8.0 ± 0.2 Buprenorphine (P* = 0.02)

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than β-endorphin are highlighted in red (One way ANOVA, followed by Students T-test, corrected for multiple comparisons, P < 0.05). * Emax or pEC50 significantly different between MOPr-WT and MOPr-N40D (P < 0.05).

160 Table 2: Summary of opioid efficacy and potency in assays of ERK1/2 phosphorylation in CHO cells expressing MOPr-WT and MOPr-N40D.

pERK1/2 Emax (%) pEC50 Opioid WT N40D WT N40D Endomorphin-2 109 ± 14 107 ± 14 8.2 ± 0.2 8.3 ± 0.1 Met-Enkephalin 103 ± 9 87 ± 6 8.1 ± 0.1 8.5 ± 0.3 Oxycodone 101 ± 8 109 ± 6 6.5 ± 0.2 6.1 ± 0.1 Fentanyl 99 ± 11 89 ± 8 8.1 ± 0.1 8.4 ± 0.2 Endomorphin-1 93 ± 6 92 ± 5 8.4 ± 0.3 8.2 ± 0.1 Methadone 88 ± 9 72 ± 8 7.3 ± 0.1 7.5 ± 0.1 -Endorphin 72 ± 9 79 ± 9 7.5 ± 0.1 7.6 ± 0.1 DAMGO 67 ± 6 72 ± 5 8.5 ± 0.1 8.5 ± 0.7 Morphine 65 ± 7 76 ± 7 7.0 ± 0.1 7.2 ± 0.1 Pentazocine 39 ± 12 39 ± 15 6.1 ± 0.2 6.0 ± 0.6 35 ± 7 14 ± 6* 8.7 ± 0.4 9.3 ± 0.1 Buprenorphine (P* = 0.04)

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than endomorphin-2 are highlighted in red (One way ANOVA, followed by Students T-test, corrected for multiple comparisons, P < 0.05). * Emax or pEC50 significantly different between MOPr- WT and MOPr-N40D (P < 0.05).

161 Opioid-mediated phosphorylation of ERK1/2

We next examined whether the N40D amino acid substitution leads to alterations in signalling through another important pathway activated by MOPr, stimulation of ERK1/2 phosphorylation by MOPr. We analysed the ability of opioids to stimulate ERK1/2 phosphorylation via MOPr-WT and MOPr-N40D in CHO cells, using a whole-cell ELISA assay.

Opioid responses were normalised against 100 nM PMA applied for 10 minutes. The average PMA response was similar in cells expressing MOPr-WT or MOPr-N40D, with corrected absorbance readings of 0.56 ± 0.07 and 0.60 ± 0.07, respectively (P > 0.5).

DAMGO-stimulated ERK1/2 phosphorylation was similar between WT and N40D expressing cells, with a pEC50 of 8.5 ± 0.1 and Emax of 67 ± 6% of the PMA response in

CHO-MOPr-WT cells and a pEC50 of 8.5 ± 0.7 and Emax of 72 ± 5% in CHO-MOPr-

N40D (Figure 2, Table 2). Preincubation of cells with 1 µM naloxone for 5 mins blocked the response for all opioids tested. Application of DAMGO to cells treated overnight with

200ng/mL PTX, or in cells where receptor expression had not been induced did not elicit a response.

The ERK1/2 phosphorylation elicited by -endorphin stimulation of MOPr-N40D was similar to that of MOPr-WT. -endorphin Emax and pEC50 were 72 ± 2% and 7.5 ± 0.1 respectively in CHO-MOPr-WT cells, and 79 ± 9% and 7.6 ± 0.1 respectively in CHO-

MOPr-N40D cells (P > 0.05). Similarly, no differences in potency or efficacy were observed between MOPr variants for endomorphin-1, endomorphin-2 and met-enkephalin

(Figure 3, Table 2).

162 Buprenorphine efficacy was significantly decreased at the N40D variant. Maximum buprenorphine stimulated ERK1/2 phosphorylation in CHO-MOPr-WT was 35 ± 7%, while in CHO-MOPr-N40D efficacy was 14 ± 6%, a reduction of approximately 60% (P <

0.05). Buprenorphine potency did not differ between cells expressing MOPr-WT and

MOPr-N40D, with pEC50s of 8.7 ± 0.4 and 9.3 ± 0.1 respectively (Figure 4, Table 2). The partial agonists morphine and pentazocine stimulated ERK1/2 phosphorylation similarly at

MOPr-WT and MOPr-N40D (Figure 4, Table 2). Fentanyl-, methadone- and oxycodone- stimulated ERK1/2 phosphorylation was also similar in efficacy and potency between

CHO-MOPr-WT and CHO-MOPr-N40D cells (Figure 5, Table 2).

There were minor differences in relative agonist efficacy between MOPr-WT and MOPr-

N40D. In assays of AC inhibition, endomorphin-1, endormorphin-2, pentazocine and buprenorphine had a significantly lower Emax than methadone for both MOPr-WT and

MOPr-N40D. DAMGO had a significantly lower Emax than methadone to inhibit AC in

MOPr-WT but not MOPr-N40D. In assays of ERK1/2 phosphorylation in both MOPr-WT and MOPr-N40D, DAMGO, -endorphin, pentazocine and buprenorphine had significantly lower Emax than the most efficacious agonist in this assay, endomorphin-2.

Morphine had a significantly lower Emax than endomorphin 2 to stimulate ERK1/2 phosphorylation in MOPr-WT but not MOPr-N40D. As there was no statistically significant difference between variants for Emax of methadone, endomorphin-2, DAMGO or morphine in any of our assays, presumably the variance inherent in the AC inhibition and ERK1/2 phosphorylation assays precluded the statistical differentiation of small differences in efficacy between ligands.

163 GIRK channel activation in AtT-20 cells

Activation of MOPr results in activation of GIRK channels via G subunits. As CHO-K1 cells do not express endogenous GIRK channels, we assessed opioid mediated GIRK activation in AtT-20 cells stably transfected with hMOPr-WT and hMOPr-N40D as previously described (Knapman et al., 2013). In AtT-20 cells loaded with membrane potential sensitive dye, application of opioids resulted in a concentration-dependent decrease in fluorescence from baseline, corresponding to membrane hyperpolarisation from GIRK activation (Figure 6).

RFU (% baseline)

100

80

60 HBSS 3nM DAMGO 40 3 mMBuprenorphine 300 nM DAMGO 20

0 060 120 180 240 300 360 Time (sec)

Figure 6: DAMGO causes membrane hyperpolarisation in AtT-20 cells expressing MOPr-WT. Raw trace showing decrease in fluorescent signal following application of vehicle (Hank’s Balanced Salt Solution, HBSS), 3 nM DAMGO, 300 nM DAMGO or 3 µM buprenorphine to AtT20 cells expressing MOPr-WT, corresponding to membrane hyperpolarisation from GIRK activation. The Y-axis is raw fluorescent units (RFU). Drugs were added for the duration of the bar. The traces are representative of at least 6 individual experiments, each performed in duplicate.

164 Sub-maximal membrane hyperpolarisation produced by low efficacy agonists such as buprenorphine or low concentrations of high efficacy agonists such as DAMGO rapidly returned towards baseline within 2 mins after addition, however maximum membrane hyperpolarisation produced by high concentrations of high efficacy agonists remained steady over 5 min. (Figure 6). There was no difference in maximum DAMGO-stimulated

GIRK activation between AtT20-MOPr-WT cells and AtT20-MOPr-N40D cells, with Emax of 34 ± 0.1% and 32 ± 0.1% decrease from baseline, respectively (P > 0.05), and pEC50 for both variants was 8.4 ± 0.01 (Figure 7, Table 3).

Buprenorphine was less potent for GIRK activation in AtT20-MOPr-N40D cells, with pEC50 of 6.7 ± 0.08 compared with pEC50 of 7.0 ± 0.07 in AtT20-MOPr-WT cells (P <

0.05). Buprenorphine efficacy was unaffected by the N40D variant, MOPr-WT Emax was

22 ± 2%, and MOPr-N40D Emax was 20 ± 2% (P > 0.05). We also observed a significant decrease in the efficacy of the partial agonist pentazocine at MOPr-N40D, the pentazocine

Emax was 8 ± 1% at MOPr-WT, and 4 ± 1% at N40D although the pEC50 was similar between variants (Figure 7, Table 3). Morphine, methadone and -endorphin signalling was unaffected at the N40D variant (Figure 7, Table 3). As buprenorphine efficacy was similar between variants, and significantly less than higher efficacy agonists, it is unlikely that the difference in pentazocine efficacy is due to lower receptor expression in AtT20-

MOPr-N40D cells. Similarly, the partial agonist activity of buprenorphine precludes the possibility of a change in receptor reserve contributing to decreased potency for GIRK activation.

165

Figure 7: Buprenorphine is less potent for GIRK activation in AtT20 cells expressing MOPr-N40D. GIRK activation was determined as described in the Methods. A) Buprenorphine activated GIRK channels to a similar degree in AtT20 cells expressing

MOPr-WT and MOPr-N40D, but with threefold lower pEC50, from 7.0 ± 0.1 in AtT20- MOPr-WT to 6.7 ± 0.1 in AtT20-MOPr-N40D (P < 0.05). B) Pentazocine activated GIRK channels with 50% lower efficacy at MOPr-N40D, Emax was decreased from 8 ± 1% in AtT20-MOPr-WT to 4 ± 1% in AtT20-MOPr-N40D (P < 0.05). C) DAMGO, morphine, - endorphin and methadone activated GIRK channels to a similar degree and with similar potency in AtT-20 cells expressing MOPr-WT and MOPr-N40D (P > 0.05). Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

166 A B

D Fluorescence (%) D Fluorescence (%) 50 50 WT WT 40 40 N40D N40D 30 30

20 20

10 10

0 0 -8 -7 -6 -5 -8 -7 -6 [Buprenorphine] log M [Pentazocine] Log M

C

D Fluorescence (%) D Fluorescence (%) 50 50 WT WT 40 40 N40D N40D 30 30

20 20

10 10

0 0 -9 -8 -7 -6 -8 -7 -6 -5 [DAMGO] log M [Morphine] log M

D Fluorescence (%) D Fluorescence (%) 50 50 WT WT 40 40 N40D N40D 30 30

20 20

10 10

0 0 -8 -7 -6 -5 -8 -7 -6 -5 [b-Endorphin] log M [Methadone] log M

167

Table 3: Summary of opioid efficacy and potency in assays of GIRK activation in AtT-20 cells expressing MOPr-WT and MOPr-N40D.

GIRK activation Emax (%) pEC50 Opioid WT N40D WT N40D DAMGO 34 ± 1 33 ± 1 8.4 ± 0.1 8.4 ± 0.1 β-Endorphin 33 ± 2 34 ± 1 6.7 ± 0.1 6.9 ± 0.1 Methadone 33 ± 2 33 ± 2 7.3 ± 0.1 7.3 ± 0.1 Morphine 31 ± 1 30 ± 1 7.6 ± 0.1 7.4 ± 0.1 22 ± 1 20 ± 2 7.0 ± 0.1 6.7 ± 0.1* Buprenorphine (P = 0.04)

8 ± 1 4 ± 1* 6.3 ± 0.6 7.0 ± 0.2 Pentazocine (P *= 0.002)

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than DAMGO are highlighted in red (One way ANOVA, followed by Students T-test, corrected for multiple comparisons, P < 0.05). Emax or pEC50 significantly different between MOPr-WT and MOPr-N40D are marked with * (P < 0.05, uncorrected for multiple comparisons).

168 DISCUSSION and CONCLUSIONS

We have shown that the commonly occurring MOPr variant N40D is less effectively activated by buprenorphine in assays of AC inhibition, ERK phosphorylation and GIRK activation. Buprenorphine efficacy was reduced by over 50% for adenylyl cyclase inhibition and ERK1/2 phosphorylation in CHO-MOPr cells, with no effect on potency.

Buprenorphine efficacy for GIRK activation was not affected by the N40D variant when expressed in AtT-20 cells, but potency was decreased threefold. In assays of GIRK activation, pentazocine efficacy was also significantly decreased, with no change in potency. No other opioids were affected by the N40D variant in any of our assays.

Buprenorphine is a partial agonist, and as such, even modest differences in levels of receptor expression between variants and/or the presence of spare receptors could have a significant impact on its signalling profile. Studies in animal models, post-mortem human brain and heterologous expression systems have suggested the N40D variant causes decreased receptor expression, although reports are inconsistent (Kroslak et al., 2007;

Mague et al., 2009; Oertel et al., 2012; Zhang et al., 2005). MOPr expression was similar for both variants in the cell lines used in our assays. Radioligand binding assays showed that CHO-MOPr-N40D cells had a slightly higher cell surface receptor expression level than CHO-MOPr-WT cells, and thus receptor expression levels are unlikely to have contributed to a decrease in buprenorphine efficacy at N40D. The N40D variant causes the removal of a putative N-linked glycosylation site, which has been shown to result in decreased MOPr glycosylation and stability in CHO cells (Huang et al., 2012). In our study, MOPr expression in CHO cells was acutely induced prior to experiments, minimizing possible effects on receptor turnover. For AtT-20 cells, receptor expression was not directly measured, however the use of the FlpIn system for receptor transfection ensures that the receptor construct is inserted only once into the same location in the

169 genome, thus all cells are subjected to a similar transcriptional environment (Sauer, 1994).

Additionally, buprenorphine efficacy was similar in cells expressing MOPr-WT and

MOPr-N40D, suggesting similar expression levels.

Our findings suggest that the amino acid change on the N-terminus of MOPr may affect the ability of some opioid ligands to effectively transduce signals to the intracellular effectors of MOPr. GPCRs constantly oscillate through a range of possible conformations, and different ligands can more effectively stabilize a subset of receptor conformations

(Kenakin & Miller, 2010). GPCRs undergo further conformational changes upon ligand binding, and the resulting conformation of the ligand-receptor complex may couple differentially to associated G-proteins (Pineyro et al., 2007). The involvement of the N- terminal domain in MOPr conformational changes is not yet well understood, and this region of the receptor is not included in published crystal structures (Manglik et al., 2012).

However, activation-dependent changes in this region have been observed in MOPr and other GPCRs. Antibodies generated against the region proximal to N40D on the N- terminal domain of MOPr show differential recognition of activated receptors, suggesting this region undergoes conformational changes upon ligand-binding (Gupta et al., 2007,

Gupta et al., 2008). This effect was not seen in cells expressing MOPr-N40D or in cells treated with deglycosylating agents, indicating that agonist binding induces movement of

N-glycan chains (Gupta et al., 2008). Furthermore, N-terminal antibody binding was affected by changes in the C-terminal region of MOPr, implying that conformational changes associated with receptor coupling to associated G-proteins may influence all domains of MOPr. A similar effect was seen in other GPCRs examined in this study including the -opioid receptor and the cannabinoid CB1 receptor (Gupta et al., 2007). N- terminal polymorphisms affect signalling in other related GPCRs. N-terminal SNPs in the

5-HT2B receptor increased constitutive and agonist-stimulated activity in COS-7 cells

170 (Belmer et al. 2014). Likewise, polymorphisms of the N-terminal region in the melanocortin-4 receptor increased constitutive activity (Srinivasan et al., 2004). Taken together, these results indicate the N-terminal region can undergo substantial structural changes upon receptor activation, and that structural changes arising from genetic variation have the ability to significantly affect receptor function. Thus, the N40D substitution may affect the ability of a ligand to bind to MOPr, or affect the efficacy of ligand-directed

MOPr coupling to effector pathways due to altered receptor conformation.

It is not immediately apparent why buprenorphine efficacy is affected in CHO cells while buprenorphine potency changes in AtT-20 cells, but the differences may be due to different effectors. MOPr inhibition of AC is mediated by Gi/o subunits of the G-protein heterotrimer, whereas GIRK activation occurs via G subunits (Law et al., 2000).

Phosphorylation of ERK1/2 can occur via multiple pathways, and may involve G and

G-coupled processes as well as G-protein independent pathways such as -arrestin

(Luttrell et al., 2005). Whereas AC inhibition is mediated by a single G subunit, GIRK channels require the binding of 4 G subunits for activation (Corey and Clapham, 2001).

Differences in the ability of the N40D variant to couple to various G-proteins may differentially affect opioid ligand signalling at G and G-mediated pathways (Allouche et al., 1999; Galandrin et al., 2008). Furthermore, cell lines vary in the available pool of G- proteins, effector molecules and regulatory proteins, and opioids may activate a different suite of G-proteins in CHO cells and AtT-20 cells (Atwood et al., 2011). Alternatively, buprenorphine may show functional selectivity towards GIRK activation over AC inhibition or ERK1/2 phosphorylation, although studies of AC inhibition and ERK1/2 phosphorylation in AtT-20 cells are necessary to determine the presence of any functional selectivity.

171 The N40D variant did not affect signalling of any of the other 10 opioid ligands tested, including -endorphin. The first in vitro studies of N40D signalling reported a threefold increase in -endorphin affinity for MOPr-N40D in AV-12 cells, as well as a threefold increase in -endorphin potency for GIRK activation in Xenopus oocytes (Bond et al.,

1998). We found no difference for MOPr-N40D in GIRK activation in AtT20 cells, or in

AC inhibition or ERK1/2 phosphorylation in CHO cells. Other studies have also failed to find differences in -endorphin coupling to other signalling pathways in various expression systems (Befort et al., 2001; Beyer et al.; 2004; Kroslak et al., 2007). Several studies have investigated the effect of N40D on ICa channel inhibition, with reports of enhanced

DAMGO and morphine signalling (Lopez Soto & Raingo, 2012; Margas et al., 2007) via

N40D, no change in DAMGO signalling (Ramchandari et al., 2011), and decreased morphine potency at N40D (Mahmoud et al., 2011). We found no significant differences in DAMGO or morphine signalling between WT-MOPr and N40D-MOPr. Most of the functional studies performed to date have failed to take into account receptor expression and/or receptor reserve, which may contribute to the inconsistency between results.

Furthermore, most of the assays were performed under different experimental conditions, making direct comparison between studies difficult (reviewed in Knapman & Connor,

2014). In our study, the assays of AC and pERK1/2 were performed on MOPr-WT and

MOPr-N40D expressing CHO cells with an isogenic cellular background, similar levels of receptor expression and in very similar experimental conditions, enabling direct comparisons of opioid potency and efficacy to be made between the variants and signalling pathways.

There are no functional studies investigating the consequences of the N40D variant on buprenorphine signalling published. However, the consistent loss of buprenorphine

172 effectiveness in several cellular expression systems and distinct signalling pathways suggest that the negative effect of the N40D variant on buprenorphine signalling may be significant in other cell types such as neurons, where MOPr is endogenously expressed. In populations where the N40D variant occurs at a high frequency (Mura et al., 2013), a significant number of people will be homozygous for the 118G allele, however the majority of carriers will be heterozygous for the 118A allele. Few clinical studies distinguish between homo- and heterozygous carriers (e.g. Bond et al., 1998; Borstav et al.¸2012; Solak et al., 2014; Song et al., 2013), and for the few studies that consider both genotypes the data is often contradictory and sample sizes are small (Klepstad et al., 2004;

Reyes-Gibby et al., 2007). There is still considerable uncertainty about the formation of obligate dimers by MOPr and other Class A GPCR in native tissue (Malik et al., 2013;

Herrick-Davis 2013), and cells expressing both N40 and D40 receptors could have 2 populations of receptors independently signalling as monomers or homodimers, or variant receptors could interact and form heterodimers. Functional studies in cells co-expressing both WT and variant receptors have not been performed, largely due to the technical difficulty of accurately quantifying and expressing equivalent amounts of each variant. In one study, HA-tagged MOPr-WT and a FLAG-tagged MOPr-L85I variant that undergoes endocytosis in response to morphine were co-expressed in HEK293 cells. Both MOPr-WT and MOPr-L85I internalised in response to morphine, suggesting the formation of functional heterodimers with L85I showing a dominant phenotype (Ravindranathan et al.,

2009), although alternative explanations are also possible. Conversely, when MOPr-WT and a non-functional MOPr-R181C variant were co-expressed, MOPr-WT internalised independently in response to DAMGO, indicating that this variant either fails to form dimers or forms unstable dimers with MOPr-WT (Ravindranathan et al., 2009). These results suggest that the potential for interaction between MOPr variants may be dependent

173 on the SNP, and the receptor dynamics resulting from co-expression of MOPr-WT and

MOPr-N40D warrant further investigation.

Buprenorphine is commonly prescribed as both an analgesic and as an alternative to methadone for the treatment of opioid dependence, and differences in signalling arising from the N40D variant may be of clinical significance (Virk et al., 2009). Most studies investigating differences in opioid requirements for N40D carriers in a clinical setting have not included buprenorphine. For example, a recent large study of European cancer patients investigated the association between the N40D variant and the required dose of opioid analgesics (Klepstad et al., 2011). This study found no association between the N40D variant and dose of morphine, fentanyl or oxycodone used, entirely consistent with our results showing no difference in the activity of these ligands at the N40D variant (Klepstad et al., 2011). Importantly, this study did not examine the interaction between OPRM1 variants and buprenorphine, and our data suggests that this would be worthwhile. One study investigating buprenorphine in N40D carriers reported a shorter opioid withdrawal period for newborns with the N40D variant who were exposed to maternal buprenorphine in utero, supporting the possibility of a reduced response of N40D carriers to buprenorphine (Wachman et al., 2013). Different individual responses to opiate maintenance treatment including buprenorphine are frequently observed among heroin addicts, however attempts to predict responders and non-responders have been unsuccessful, and the effect of the N40D variant has not been investigated (Gerra et al.,

2014).

Our results demonstrate that the N40D variant has the potential to affect MOPr function across cell types and signalling pathways. Further functional studies examining other

MOPr signalling pathways are required to extend our findings, and investigation into

174 buprenorphine response in N40D carriers in a clinical setting would be of great interest.

The N40D variant is highly prevalent within the population, occurring at allelic frequencies ranging from 10 – 50% (Mura et al., 2013). A decrease in buprenorphine efficacy arising from the N40D variant could be a contributing factor for the lack of response of some individuals to buprenorphine maintenance therapy, and may result in a significant proportion of the population receiving inadequate or inappropriate analgesic therapy. Although individual opioid response is influenced by a number of genetic and epigenetic factors, understanding how the N40D variant affects cellular signalling is a key element in predicting the potential clinical or phenotypic consequences of a particular opioid drug. Prior knowledge of individual genotype could provide valuable insight into the most effective form of opioid therapy, minimizing the risk of serious adverse events associated with opioid overdose, while maximizing therapeutic benefits and ensuring individuals receive adequate pain relief.

Acknowledgements:

This study was supported the National Health and Medical Research Council of Australia project grant 1011979 to MC. AK and MS were recipients of Postgraduate Scholarships from Macquarie University, topped up by the Australian School of Advanced Medicine.

We thank Mac Christie and Yan Ping Du for their assistance with the binding assays, and

Courtney Breen for providing a protocol and advice on the ERK ELISA.

Author Contributions

AK, MS and MC designed and analysed experiments, AK and MS conducted the experiments. AK and MC conceived the study and wrote the paper, all authors have seen the final manuscript.

175 References

Abrol R, Kim SK, Bray JK, Trzaskowski B, Goddard WA 3rd (2013). Conformational ensemble view of G protein-coupled receptors and the effect of mutations and ligand binding. Methods Enzymol 520: 31-48.

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al.,

(2013). The concise guide to PHARMACOLOGY 2013/14: G protein-coupled receptors.

Brit J Pharmacol 170: 1459-1581.

Atwood BK, Lopez J, Wager-Miller J, Mackie K, Straiker A (2011). Expression G protein- coupled receptors and related proteins in HEK293, AtT20, BV2, and N18 cell lines as revealed by microarray analysis. BMC Genomics 12:14

Allouche S, Polastron J, Hasbi A, Homburger V, Jauzac P (1999). Differential G-protein activation by alkaloid and peptide opioid agonists in the human neuroblastoma cell line

SK-N-BE. Biochem J 342: 71-78.

Belmer A, Doly S, Setola V, Banas SM, Moutkine I, Boutourlinsky K et al. (2014). Role of the N-terminal region in G protein-coupled receptor functions: Negative modulation revealed by 5-HT2B receptor polymorphisms. Mol Pharmacol 85: 127-138.

Befort K, Filliol D, Decaillot FM, Gaveriaux-Ruff C, Hoehe MR, Kieffer BL (2001). A single nucleotide polymorphic mutation in the human mu-opioid receptor severely impairs receptor signalling. J Biol Chem 276: 3130-7.

176 Beyer A, Koch T, Schroder H, Schulz S, Hollt V (2004). Effect of the A118G polymorphism on binding affinity, potency and agonist-mediated endocytosis, desensitisation and resensitisation of the human mu-opioid receptor. J Neurochem 89: 553-

560.

Bond C, LaForge KS, Tian M, Melia D, Zhang S, Borg L et al. (1998). Single-nucleotide polymorphism in the human mu opioid receptor gene alters -endorphin binding and activity: Possible implications for opiate addiction. Proc Natl Acad Sci USA 95: 9608-

9613.

Borstov AV, Millikan RC, Belfer I, Boortz-Marx RL, Arora H, McLean SA (2012). µ-

Opioid receptor gene A118G polymorphism predicts survival in patients with breast cancer. Anesthesiology 116: 896-902.

Corey S, Clapham DE (2001). The stoichiometry of G binding to G-protein-regulated inwardly rectifying K+ channels (GIRKs). J Biol Chem 276: 11409-11413.

Cox B (2013). Recent developments in the study of opioid receptors. Mol Pharmacol

83:723-728.

Cox BM, Christie MJ, Devi L, Toll L, Traynor JR (2014). Challenges for opioid receptor nomenclature. Brit J Pharmacol, In Press

Diatchenko L, Robinson JE, Maixner W (2011). Elucidation of mu-opioid gene structure:

How genetics can help predict therapeutic response to opioids. Eur J Pain Suppl 5: 433-48.

177 Galandrin S, Oligny-Longpre G, Bonin H, Ogawa K, Gales C, Bouvier M (2008).

Conformational rearrangements and signalling cascades involved in ligand-biased mitogen-activated protein kinase signalling through the 1-adrenergic receptor. Mol

Pharmacol 74: 162-172.

Gerra G, Somaini L, Leonardi C, Cortese E, Maremmani I, Manfredini M, et al., (2014).

Association between gene variants and response to buprenorphine maintenance treatment.

Psychiatry Res 215: 202-207

Gupta A, Decaillot FM, Gomes I, Tkalych O, Heimann AS, Ferro ES et al. (2007).

Conformation state-sensitive antibodies to G-protein-coupled receptors. J Biol Chem 282:

5116-5124.

Gupta A, Rozenfeld R, Gomes I, Raehal KM, Decaillot FM, Bohn LM et al. (2008). Post- activation-mediated changes in opioid receptors detected by N-terminal antibodies. J Biol

Chem 283: 10735-10744.

Herrick-Davis K, Grinde E, Cowan A, Mazurkiewicz JE (2013). Fluorescence correlation spectroscopy analysis of serotonin, adrenergic, muscarinic, and dimerisation: the oligomer number puzzle. Mol Pharmacol 84: 630-642.

Huang P, Chen C, Mague SD, Blendy JA, Liu-Chen LY (2012). A common single nucleotide polymorphism A118G of the µ opioid receptor alters its N-glycosylation and protein stability. Biochem J 441: 379-386.

178 Kenakin T, Miller LJ (2010). Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery.

Pharmacol Rev 62: 265-304.

Klepstad P, Fladvad T, Skorpen F, Bjordal K, Caraceni A, Dale O et al. (2011). Influence from genetic variability on opioid use for cancer pain: A European genetic association study of 2294 cancer pain patients. Pain 152: 1139-1145.

Klepstad P, Rakvag TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC et al. (2004). The

118 A > G polymorphism in the human µ-opioid receptor gene may increase morphine requirements in patients with pain caused by malignant disease. Acta Anaesthesiol Scand

48: 1232-1239.

Knapman A, Abogadie F, McIntyre P, Connor M (2014). A real-time, fluorescence-based assay for measuring µ-opioid receptor modulation of adenylyl cyclase activity in Chinese hamster ovary cells. J Biomol Screen 19: 223-231.

Knapman A, Connor M (2014). Cellular signalling of non-synonymous single nucleotide polymorphisms of the human µ-Opioid receptor (OPRM1). Br J Pharmacol, In Press doi:

10.1111/bph.12644

Knapman A, Santiago M, Du YP, Bennallack PR, Christie MJ, Connor M (2013). A continuous, fluorescence-based assay of µ-opioid receptor activation in AtT-20 cells. J

Biomol Screen 18: 269-76.

179 Kroslak T, LaForge KS, Gianotti RJ, Ho A, Nielsen DA, Kreek MJ (2007). The single nucleotide polymorphism A118G alters functional properties of the human mu opioid receptor. J Neurochem 103: 77-87.

Law PY, Yung HW, Loh HH (2000). Molecular mechanisms of opioid receptor signalling.

Annu Rev Pharmacol Toxicol 40: 389-430.

Lopez Soto EJ, Raingo J (2012). A118G mu opioid receptor polymorphism increases inhibitory effects on CaV2.2 channels. Neurosci Lett 523: 190-194.

Lotsch J, Geisslinger G (2005). Are µ-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol Med 11: 82-89.

Luttrell LM (2005). Composition and function of G protein-coupled receptor signalsomes controlling mitogen-activated protein kinase activity. J Mol Neurosci 26L 253-264.

Mague SD, Isiegas C, Huang P, Liu-Chen LY, Lerman C, Blendy JA (2009). Mouse model of OPRM1 (A118G) polymorphism has sex-specific effects on drug-mediated behavior.

Proc Natl Acad Sci USA 106: 10847-10852.

Mague SD, Blendy JA (2010). OPRM1 SNP (A118G): involvement in disease development, treatment response, and animal models. Drug Alcohol Depend 108: 172-182.

Mahmoud S, Thorsell A, Sommer WH, Heilig M, Holgate JK, Bartlett SE, Ruiz-Velasco V

(2011). Pharmacological consequence of the A118G μ opioid receptor polymorphism on

180 morphine- and fentanyl-mediated modulation of Ca2+ channels in humanized mouse sensory neurons. Anesthesiology 115: 1054-1062.

Malik RU, Ritt M, DeVree BT, Neubig RR, Sunahara RK, Sivaramakrishnan S (2013).

Detection of G protein-selective G protein-coupled receptor (GPCR) conformations in live cells. J Biol Chem 288: 17167-17178.

Manglik A, Kruse AC, Kobilka TS, Thian FS, Mathiesen JM, Sunahara RK et al. (2012).

Crystal structure of the µ-opioid receptor bound to a morphinan antagonist. Nature 485:

321-326.

Margas W, Zubkoff I, Schuler HG, Janicki PK, Ruiz-Velasco V (2007). Modulation of

Ca2+ channels by heterologously expressed wild-type and mutant human µ opioid receptors

(hMORs) containing the A118G single-nucleotide polymorphism. J Neurophysiol 97:

1058-1067.

Matsui A, Williams JT (2010). Activation of µ-opioid receptors and block of KIR3 potassium channels and NMDA receptor conductance by l- and d- methadone in rat locus coeruleus. Br J Pharmacol 161: 1403-1413.

Merikangas KR, Stolar M, Stevens DE, Goulet J, Preisig MA, Fenton B et al. (1998).

Familial transmission of substance use disorders. Arch Gen Psychiatry 55: 973-9.

Mura E, Govoni S, Racchi M, Carossa V, Ranzani GN, Allegri M et al., (2013).

Consequences of the 118A>G polymorphism in the OPRM1 gene: translation from bench to bedside? J Pain Res 6: 331-53.

181 Nguyen VH, Ingram SL, Kassiou M, Christie MJ (1998). Sigma-binding site ligands inhibit K+ currents in rat locus coeruleus neurons in vitro. Eur J Pharmacol 361: 157-163.

Oertel BG, Doehring A, Roskam B, Kettner M, Hackmann N, Ferreiros N et al. (2012).

Genetic-epigenetic interaction modulates µ-opioid receptor regulation. Hum Mol Genet 21:

4751-4760.

Pineyro G, Archer-Lahlou E (2007). Ligand-specific receptor states: implications for opiate receptor signalling and regulation. Cell Signal 19: 8-19.

Ramchandani VA, Umhau J, Pavon FJ, Ruiz-Velasco V, Margas W, Sun H et al. (2011). A genetic determinant of the striatal dopamine response to alcohol in men. Mol Psychiatry

16: 809-817.

Reyes-Gibby CC, Shete S, Rakvag T, Bhat SV, Skorpen F, Bruera E et al (2007).

Exploring joint effects of genes and the clinical efficacy of morphine for cancer pain:

OPRM1 and COMT gene. Pain 130: 25-30.

Sauer B (1994). Site-specific recombination: Developments and applications. Curr Opin

Biotechnol 5: 521-527.

Singh VK, Bajpai K, Biswas S, Haq W, Khan MY, Mathur KB (1997). Molecular biology of opioid receptors. Neuroimmunomodulation 4: 285-297.

Skorpen F, Laugsend EA (2008). Variable response to opioid treatment: any genetic predictors in sight? Palliat Med 22: 310-327.

182 Solak O, Erdogan MO, Yildiz H, Ulasli AM, Yaman F, Terzi ES et al. (2014). Assessment of opioid receptor µ1 gene A118G polymorphism and its association with pain intensity in patients with fibromyalgia. Rheumatol Int, In Press doi: 10.1007/200296-014-2995-1.

Somogyi AA, Barrat DT, Coller JK (2007) Pharmacogenetics of opioids. Clin Pharmacol

Ther 81: 429-444.

Song Z, Du B, Wang K, Shi X (2013). Effects of OPRM1 A118G polymorphism on epidural analgesia with fentanyl during labor: a meta-analysis. Genet Test Mol Biomarkers

17: 743-749.

Srinivasan S, Lubrano-Berthelier C, Govaerts C, Picard F, Santiago P, Conklin BR et al

(2004). Constitutive activity of the melanocortin-4 receptor is maintained by its N-terminal domain and plays a role in energy homeostasis in humans. J Clin Invest 114: 1158-1164.

Virk MS, Arttamangkul S, Birdsong WT, Williams JT (2009). Buprenorphine is a weak partial agonist that inhibits opioid receptor desensitisation. J Neurosci 29: 7341-7348.

Wachman EM, Hayes MJ, Brown MS, Paul J, Harvey-Wilkes K, Terrin N, et al., (2013).

Association of OPRM1 and COMT single-nucleotide polymorphisms with hospital length of stay and treatment of neonatal abstinence syndrome. JAMA 309: 1821-1827.

Walter C, Lotsch J (2009). Meta-analysis of the relevance of the OPRM1 118A>G genetic variant for pain treatment. Pain 146: 270-275.

183 Zhang Y, Wang D, Johnson AD, Papp AC, Sadee W (2005). Allelic expression imbalance of human mu opioid receptor (OPRM1) caused by variant A118G. J Biol Chem 280:

32618-24.

184 Chapter 6

The A6V polymorphism of the human µ-opioid receptor negatively

impacts signalling of morphine and endogenous opioids in vitro

Alisa Knapman, Marina Santiago and Mark Connor

This paper has been prepared for future submission to the British Journal of

Pharmacology. Marina Santiago constructed AtT20-MOPr cells and performed GIRK activation assays. Mark Connor assisted with experimental design, data analysis and manuscript preparation. All other work is my own.

185 SUMMARY

Background and Purpose

Evidence suggests that polymorphisms of the μ-opioid receptor (MOPr) may contribute to individual variation in opioid response. The A6V variant is present in up to 20% of individuals in some populations, and has been associated with increased susceptibility to substance abuse. To date, no functional studies have examined the effect of A6V on MOPr signalling in vitro. We sought to determine functional differences in MOPr signalling at several signalling pathways using a range of structurally distinct opioid ligands in cells expressing wild-type MOPr and the MOPr variant, A6V.

Experimental Approach

MOPr-WT and MOPr-A6V were stably expressed in CHO cells and in AtT-20 cells.

Assays of adenylyl cyclase inhibition and ERK1/2 phosphorylation were performed on

CHO cells, and assays of GIRK activation were performed on AtT-20 cells. Signalling profiles for each ligand were compared between variants.

Key Results

Buprenorphine efficacy was abolished at MOPr-A6V for adenylyl cyclase inhibition and

ERK1/2 phosphorylation, but not GIRK activation. DAMGO, morphine and β-endorphin signalling was significantly compromised for AC inhibition, and signalling of all opioids tested was reduced for ERK1/2 phosphorylation. For GIRK activation, morphine signalling was selectively affected by the A6V variant.

Conclusions and Implications

The A6V variant is present in 1 - 20% of the population. This variant affects the signalling of many clinically important opioids including morphine, buprenorphine and fentanyl, as well as the signalling of endogenous opioids. This may affect individual response to opioid therapy, and possible disruption of the endogenous opioid system may contribute to susceptibility to substance abuse.

186 Abbreviations

AC adenylyl cyclase

BSA bovine serum albumin

CHO Chinese Hamster Ovary

DAMGO ([D-Ala2, N-MePhe4, Gly-ol]-enkephalin)

DMEM Dulbecco’s Modified Eagle Medium

ERK extracellular signal regulated kinase

FSK forskolin

GIRK G protein-gated inwardly rectifying K channel

GPCR G protein couple receptors

MOPr µ-opioid receptor

OPRM1 opioid receptor mu 1

PTX pertussis toxin

187 INTRODUCTION

Morphine and other opioid analgesics are the most effective treatments currently available for moderate to severe pain. In addition to their powerful analgesic effects, opioids are associated with a number of negative side effects including respiratory depression, nausea, constipation and sedation, as well as the development of analgesic tolerance with continued use (Boom et al., 2012; Matthes et al., 1996). Opioids are characterised as having a narrow therapeutic window, and for most opioids there is a fine balance in dosing to achieve a therapeutic dose without producing dangerous side effects such as respiratory depression (Somogyi et al., 2007). Furthermore, the degree to which individuals experience opioid-induced analgesia and adverse effects is highly variable and difficult to predict (Lotsch & Geisslinger, 2005; Merikangas et al., 2008). This can lead to inadequate pain relief for many individuals, as dosing of opioid analgesics may be restricted in order to avoid serious adverse events (Skorpen & Laugsand, 2008).

Individual differences in opioid requirements and pain perception are likely to have a genetic basis (Crist & Berrettini, 2013). The µ-opioid receptor (MOPr) is the primary target for most clinically prescribed opioid analgesics including morphine, buprenorphine and oxycodone (Matthes et al., 1996; Nozaki & Komei, 2007; Virk et al., 2009). A number of naturally occurring, non-synonymous genetic variants have been identified in the coding region of the MOPr gene OPRM1, and there is evidence to suggest that some of these may affect individual opioid response (Lotsch & Geisslinger, 2005; Somogyi et al., 2007). The two most common OPRM1 variants are the N40D variant and the A6V variant, both single nucleotide polymorphisms (SNPs) in the N-terminal region of MOPr. The N40D variant, present at allelic frequencies ranging from 10-50% in various populations, has been studied extensively and associated with a diverse range of clinical outcomes, including differences in pain perception, opioid requirements, and predisposition to substance abuse, however

188 these associations have not been consistently found (Diatchenko et al., 2011; Klepstad et al., 2011, Mague & Blendy, 2010; Mura et al., 2013; Walter & Lotsch, 2009). Likewise, functional studies of MOPr show differences between N40D and wild-type (WT) receptor signalling, but with little consistency in results (Knapman & Connor, 2014).

Despite being reported at allelic frequencies upwards of 20% in African-American and northern Indian populations (e.g. Hoehe et al., 2000; Crowley et al., 2003; Kapur et al.,

2007, Crystal 2010), the A6V variant has received far less attention than the N40D variant.

Data suggest that there may be a higher frequency of the V6 allele in substance-abusing populations (Berrettini et al., 1997; Compton et al., 2003; Crowley et al., 2003; Crystal et al., 2010; Rommelspacher et al., 2001), but the effects of the polymorphism on analgesic responses is unknown. Only two studies have examined the functional consequences of the

A6V variant on MOPr signalling in vitro. In an assay of cAMP-dependent gene transcription in HEK-293 cells transiently expressing MOPr, there was no difference in the potency of DAMGO, endomorphin-1 or leu-enkephalin between the A6 and V6 variants

(Fortin et al., 2010). Another study expressed the V6 variant on the MOPr-1A splice variant backbone in HEK-293 cells co-transfected transiently with a chimeric G protein

(Ravinadrathan et al., 2009). In this study, the maximal effect of DAMGO but not morphine in mediating intracellular Ca release was increased at the MOPr-1A-V6 variant compared with MOPr-1A-A6. These studies provide little insight into how the MOPr-A6V polymorphism affects acute MOPr function, so in this study, we investigated the ability of human MOPr-WT and MOPr-A6V receptors to couple to several different signalling pathways in CHO cells and AtT-20 cells using 11 clinically important and/or structurally distinct opioid ligands. We found that the substitution of valine for alanine at amino acid 6 of MOPr had a significant, negative impact on the ability of the receptor to inhibit adenylyl cyclase (AC) and to stimulate extracellular signal-regulated kinase 1/2 (ERK1/2)

189 phosphorylation in CHO cells. Notably, buprenorphine signalling was abolished in CHO-

MOPr-A6V cells for both AC inhibition and ERK1/2 phosphorylation. By contrast, the effects of the A6V substitution on activation of G protein-gated, inwardly rectifying potassium channels (GIRK, composed of KIR 3.x subunits) were very modest, with only morphine showing any reduction in efficacy. These data suggest that the common A6V polymorphism of MOPr results in a receptor that may show pathway specific changes in function.

METHODS

MOPr transfection and cell culture

CHO-FRT-TRex cells were stably transfected with a pcDNA5 construct encoding the haemagglutinin-tagged human -opioid receptor cDNA together with the pOG44 (Flp recombinase plasmid) using the transfectant Fugene (Promega), as described previously

(Knapman et al., 2014). The HA-tagged human wild-type -opioid receptor (MOPr-WT) and the -opioid receptor containing the V6 variant (MOPr-A6V) were synthesised by

Genscript (Piscataway, New Jersey, USA). Cells expressing MOPr-WT or MOPr-A6V were selected using hygromycin B (500 µg/mL). Cells were cultured in Dulbecco’s

Modified Eagle Medium (DMEM) containing 10% FBS, 100U penicillin/streptomycin and

500 g/mL hygromycin B up to passage 5. Hygromycin concentration was reduced to 200

g/mL beyond passage 5.

AtT-20 FlpIn cells were constructed by transfecting Flp-recombinase target site

(FRT)/LacZeo2 (Life Technologies) using Fugene. Successfully transfected cells were selected with 100 µg/mL zeocin, and stably transfected with MOPr-WT or MOPr-A6V as described for CHO cells. Cells expressing MOPr were selected using 100 µg/mL

190 hygromycin B. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% FBS, 100U penicillin/streptomycin and 100 g/mL hygromycin B.

CHO-MOPr cells and AtT-20-MOPr cells were passaged at 80% confluency as required.

Assays were carried out on cells up to 30 passages. Cells for assays were grown in 75 cm2 flasks and used at greater than 90% confluence.

MOPr receptor expression

MOPr surface expression was determined on CHO-MOPr cells by incubating intact cells with 0.125 – 16 nM [3H]-DAMGO (D-Ala2, N-MePhe4, Gly-ol]-enkephalin; PerkinElmer,

Waltham, MA) for 2 h at 4 °C in 50 mM Tris-Cl (pH 7.4). Briefly, 24 hours before the assay, cells were detached from flasks with trypsin/EDTA (Sigma) and resuspended in

DMEM containing 10% FBS, 100U penicillin/streptomycin, plus 2 µg/mL tetracycline to induce MOPr expression. Cells were seeded at a density of 2 × 105 cells per well in 24- well plates pre-coated with polylysine and grown overnight at 37° C. On the day of the assay, cells were washed twice gently with 50 mM Tris-Cl (pH 7.4) and incubated for 2 h on ice with 0.125 – 16 nM [3H]-DAMGO. Non-specific binding was determined in the presence of unlabeled DAMGO (10 μM). Non-specific binding was 15 ± 1% of total binding for CHO-MOPr-WT, and 12 ± 1% in CHO-MOPr-A6V. At the end of the incubation, cells were washed three times with 50 mM Tris-Cl (pH 7.4) at 4 °C. Cells in each well were then digested for 1 h at room temperature with 100 μL of 1N NaOH. 100

μL 1N HCl was added to each well to stop the reaction and cells were then collected into scintillation vials. Total bound [3H]-DAMGO was determined using a liquid scintillation counter (Packard Tricarb, Perkin Elmer,Waltham MA, USA). Receptor density (Bmax) and affinity (Kd) were calculated using a one-site binding curve fitted using GraphPad Prism

(GraphPad Software, La Jolla, CA). Protein concentration per well was determined with a

191 BCA Protein Assay Kit (Pierce) according to the manufacturer’s instructions. All experiments were performed three times in triplicate.

Membrane Potential Assay of Adenylyl Cyclase Inhibition

Opioid inhibition of adenylyl cyclase was measured using a membrane potential assay as previously described (Knapman et al., 2014). Briefly, 24 hrs prior to the assay, CHO-

MOPr cells were detached from the flask with trypsin/EDTA (Sigma) and resuspended in

10 ml of Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin and 15 mM glucose. MOPr receptor expression was induced with the addition of 2 g/mL tetracycline. The cells were plated in a volume of 90 µl per well in black walled, clear-bottomed 96-well microplates (Corning), and incubated overnight at

o 37 C in ambient CO2. Membrane potential was measured using a FLIPR Membrane

Potential Assay kit (blue) from Molecular Devices. The dye was reconstituted with assay buffer containing (in mM), NaCl 145, HEPES 22, Na2HPO4 0.338, NaHCO3 4.17, KH2PO4

0.441, MgSO4 0.407, MgCl2 0.493, CaCl2 1.26, glucose 5.56 (pH 7.4, osmolarity 315 ± 5).

Prior to the assay, cells were loaded with 90 l/well of the dye solution without removal of the L-15 and incubated at 37°C for 60 minutes in ambient CO2. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at 565 nm. Baseline readings were taken every 2 seconds for at least 2 minutes, at which time FSK + opioid or FSK + vehicle was added in a volume of 20 l. Further additions were made in volumes of 20 µl, as indicated.

The background fluorescence of cells without dye or dye without cells was negligible.

Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle addition. When used, the final

192 concentration of the solvents DMSO or EtOH was not more than 0.1%, and these concentrations did not produce a signal in the assay.

Membrane Potential Assay of GIRK Channel Activation

Opioid activation of endogenous GIRK channels in AtT-20 cells was measured using a membrane potential sensitive dye, as previously described (Knapman et al., 2013). AtT-

20-MOPr cells were detached from flasks and plated using the same procedure as for the

CHO-MOPr cells in the AC inhibition assay, without the addition of tetracycline. Blue membrane potential dye was reconstituted and loaded onto cells, and the assay was performed in the same way as the AC inhibition assay. Baseline readings were taken every

2 seconds for at least 2 minutes, at which time opioid or vehicle was added in a volume of

20 l. The background fluorescence of cells without dye or dye without cells was negligible. The final concentration of DMSO was not more than 0.1%, and this concentration did not produce a signal in the assay.

ELISA of ERK1/2 phosphorylation

Opioid-mediated ERK1/2 phosphorylation was measured by ELISA, as previously described (Knapman et al., In Press). Briefly, 24 hrs before the assay, CHO-MOPr cells were plated in 96-well clear microplates (Falcon) and receptor expression was induced as for the AC inhibition assay. Cells were serum-starved for 1 hr in 40 μL serum-free L-15 supplemented with 5% bovine serum albumin (BSA) before the assay. Cells were treated with drug or vehicle diluted in serum-free L-15 (40 µl added) with no BSA added to minimise background. Preliminary time-course experiments indicated that a 5 min drug treatment was optimal to produce robust ERK1/2 phosphorylation without inducing desensitisation, thus all drug treatments were for 5 min. After drug application, the reaction was stopped by inverting the plates to remove the drug solution, placing the plates on ice,

193 and immediately fixing the cells with 4% paraformaldehyde for 15 min at RT. Cells were washed 3 times with 300μL PBS, then permeabilized with 0.1% Triton-X in PBS for 30 min at RT. Triton-X was removed, and cells were incubated for 2 hr at RT with blocking solution consisting of 5% BSA in PBS with 0.01% Tween-20 (PBS-T). Blocking solution was removed, then cells were incubated overnight at 4°C with a 1:500 dilution of rabbit α- phospho-p44/42 MAPK (Thr202/Tyr204) antibody in PBS-T with 1% BSA. Cells were washed 3 times with 300μL PBS-T, and incubated with 1:5000 α-rabbit IgG HRP-linked antibody in PBS-T with 1% BSA for 2 hr. Cells were washed 4 times with 300 μL PBS-T, and incubated with 3,3’,5,5’-tetramethylbenzidine (TMB; Sigma) at RT in the dark for 45 min. The reaction was stopped with 1M HCl. Absorbance was read at 450 nm using a

BMG Pherastar FS microplate reader. Cells were then stained with 0.5μg/mL DAPI for 10 min at RT, and washed 3 times with 300 μL PBS-T. Fluorescence was read in the Pherastar microplate reader with cells excited at a wavelength of 358 nm and emission measured at

461 nm. Absorbance readings were normalised to DAPI staining to account for any differences in cell density between wells. Readings were then normalised to the response of cells treated with 100nM PMA for 10 min.

Drugs and Chemicals

Tissue culture reagents and buffer salts were from Invitrogen or Sigma unless otherwise noted. Tyr-D-Ala-Gly-N-MePhe-Gly-ol acetate (DAMGO), endomorphin-1, endomorphin-

2 and met-enkephalin were purchased from Auspep (Tullamarine, Australia). Morphine, fentanyl and pentazocine were a kind gift from the Department of Pharmacology,

University of Sydney. Buprenorphine and oxycodone were from the National

Measurement Institute (Lindfield, Australia). β-endorphin was from Genscript

(Piscataway, New Jersey, USA). Forskolin and naloxone were from Ascent

Pharmaceuticals (Bristol, UK). 1,9-dideoxyforskolin and tetraethylammonium (TEA) were

194 from Sigma Aldrich (Castle Hill, Australia). Pertussis toxin (PTX) was from Tocris

Bioscience (Bristol, UK). Phospho-ERK1/2 antibody (Catalog #9101) and anti-rabbit IgG

HRP-lined antibody (Catalog #7074) were from Cell Signalling Technologies, Danvers,

Massachusetts, USA.

Data

Unless otherwise noted, data is expressed as mean ± s.e.m. of at least 5 determinations made in duplicate or triplicate. Concentration response curves (CRCs) were fit with a 4 parameter logistic equation using Graphpad Prism (Graphpad). CRCs from separate experiments were pooled and compared with two-way ANOVA, followed by Bonferroni post-hoc test corrected for multiple comparisons. Emax and pEC50 values derived from individual experiments were compared using unpaired Student’s t-test. Comparisons between maximum agonist responses were made by comparing Emax values derived from individual experiments using one-way ANOVA followed by Dunnett-s post-hoc test, corrected for multiple comparisons. All other comparisons were made using an unpaired

Student’s T-test. P < 0.05 was considered significant. Channel and receptor nomenclature is consistent with the British Journal of Pharmacology Concise Guide to Pharmacology

2013/2014 and recent guidelines from IUPHAR (Alexander et al., 2013, Cox et al., 2014).

RESULTS

MOPr Expression in CHO-K1 cells

CHO-K1 cells were stably transfected with either MOPr-WT or the MOPr-A6V variant, with receptor expression controlled by a tetracycline-sensitive repressor. After 24 h tetracycline induction, surface receptor expression was similar in cells expressing MOPr-

WT or MOPr-A6V as indicated by similar levels of specific [3H]DAMGO binding (see

Figure 1). The Bmax for CHO-MOPr-WT cells was 280 ± 20 fmol/mg total protein and 301

195 ± 25 fmol/mg for CHO-MOPr-A6V (P > 0.05). The affinity for [3H]DAMGO was also similar between for MOPr-WT and MOPr-A6V expressing cells, with KD of 0.75 ± 0.10 and 0.55 ± 0.05, respectively (P > 0.05).

Specific Binding 300 (fmol/mg total protein)

200

WT 100 A6V

0 0510 15 20 [DAMGO] nM

Figure 1: Saturation binding curve of [3H]DAMGO in intact CHO-MOPr-WT and CHO-MOPr-A6V cells, 24 hr after induction of receptor expression with tetracycline. Radioligand binding was carried out as described in the Methods. No significant difference in Bmax or Kd was observed between cells expressing MOPr-WT or MOPr-A6V (P > 0.05). Each point represents the mean ± s.e.m. of triplicate determinations from a single experiment. The assay was repeated 3 times.

196 Adenylyl Cyclase Inhibtion

In CHO-MOPr cells loaded with membrane potential dye, addition of the AC activator forskolin (FSK, 300 nM) produced a rapid decrease in fluorescence, consistent with hyperpolarisation of the cells (Figure 2; Knapman et al., 2014). The hyperpolarisation stabilised 5 min after the addition of FSK, at which point measurements were taken. FSK

(300 nM) produced a similar hyperpolarisation in CHO-MOPr-WT cells (41 ± 1 % decrease in fluorescence) and CHO-MOPr-A6V cells (39 ± 2 % decrease, P > 0.05).

Simultaneous addition of the MOPr selective peptide agonist DAMGO with FSK resulted in a concentration-dependent inhibition of the FSK-stimulated hyperpolarisation in both cell lines (Figure 2). DAMGO maximally inhibited the FSK response by 62 ± 3 %, with pEC50 of 7.9 ± 0.1 in CHO-MOPr-WT cells (Table 1). We have reported previously that the effects of DAMGO in this assay were sensitive to naloxone and strongly inhibited by overnight pretreatment with pertussis toxin (Knapman et al., 2014). Application of opioids alone did not affect the membrane potential of CHO-MOPr cells, with the exception of 30

µM pentazocine and 10 µM methadone, which caused small, transient increases in fluorescence (< 10%). This effect was not naloxone sensitive.

In cells expressing MOPr-A6V, DAMGO inhibition of the FSK-induced hyperpolarisation was significantly reduced compared with cells expressing MOPr-WT, with Emax of 52 ±

3%, and pEC50 of 7.6 ± 0.2 (Table 1). A 2-way ANOVA indicated the DAMGO CRCs for

CHO-MOPr-WT and CHO-MOPr-A6V were significantly different (P < 0.001, Figure 2).

197

Figure 2: DAMGO inhibits adenylyl cyclase and activates ERK1/2 in CHO cells expressing MOPr-WT or MOPr-A6V. Adenylyl cyclase inhibition and levels of ERK1/2 phosphorylation were determined as described in the Methods. A) Raw trace showing decrease in fluorescent signal following application of 300 nM forskolin + HBSS, and 300 nM forskolin + 1 μM DAMGO at 120 sec. B) DAMGO inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation in CHO cells expressing MOPr-WT or MOPr-A6V was significantly affected by the A6V variant (2-way ANOVA, P < 0.001). C) DAMGO- stimulated ERK1/2 phosphorylation in CHO-MOPr cells was significantly affected by the A6V variant (2-way ANOVA, P < 0.0001). Maximum ERK1/2 phosphorylation by 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from at least 5-6 independent determinations performed in duplicate.

198 A

RFU (% baseline)

100

80

60

40 300 nM FSK + HBSS 300 nM FSK + 1 mMDAMGO (WT) 20 300 nM FSK + 1 mMDAMGO (A6V) 0 0120 240 360 480 600 720 840 Time (secs) B

Inhibition of FSK response 100 (%)

80 WT A6V 60

40

20

0 -9 -8 -7 -6 -5 [DAMGO] log M

C

pERK1/2 (% Control) 100

80 WT A6V 60

40

20

0 -10 -9 -8 -7 -6 [DAMGO] log M

199

Table 1: Summary of opioid efficacy and potency in assays of AC inhibition in CHO cells expressing MOPr-WT and MOPr-A6V.

AC inhibition Emax (%) pEC50 Opioid WT A6V WT A6V β-Endorphin **** 73 ± 4 50 ± 4* 6.8 ± 0.1 6.4 ± 0.1* Fentanyl ** 70 ± 6 68 ± 6 8.0 ± 0.2 7.6 ± 0.1 Morphine **** 66 ± 4 39 ± 4* 7.0 ± 0.1 6.1 ± 0.3* Oxycodone *** 65 ± 5 69 ± 8 5.8 ± 0.3 5.4 ± 0.4 Methadone **** 65 ± 3 52 ± 9 7.0 ± 0.1 6.4 ± 0.1* Met-Enkephalin ** 64 ± 4 67 ± 6 7.9 ± 0.1 7.2 ± 0.1* Endomorphin-1 *** 64 ± 5 61 ± 5 8.2 ± 0.1 7.9 ± 0.2 Endomorphin-2 **** 65 ± 7 56 ± 6 8.2 ± 0.1 7.7 ± 0.1* DAMGO *** 62 ± 3 52 ± 3* 7.9 ± 0.1 7.6 ± 0.2 38 ± 7 No 8.4 ± 0.3 No Buprenorphine Response Response Pentazocine * 36 ± 5 24 ± 7 N/A N/A

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than β-endorphin are highlighted in red (1-way ANOVA, followed by Dunnett’s post-hoc test corrected for multiple comparisons, P < 0.05). Opioids with concentration-response curves significantly different between variants are marked with * (2-way ANOVA, followed by Bonferroni post-hoc test, corrected for multiple comparisons. * P < 0.05, ** P < 0.01, *** P < 0.001,

**** P < 0.0001). MOPr-A6V Emax and pEC50 values significantly different from MOPr- WT are marked with * (unpaired Student’s T-test, P <0.05).

200 We next determined the potency and efficacy of a range of clinically important and/or structurally distinct opioid ligands in CHO cells expressing WT-MOPr or the MOPr-A6V to determine whether the A6V amino acid substitution affected MOPr inhibition of AC in a ligand-selective manner. In cells expressing MOPr-WT, morphine inhibited the FSK response by 66 ± 4%, with pEC50 of 7.0 ± 0.1. In CHO-MOPr-A6V cells, morphine inhibition of FSK effects was significantly compromised, with a pEC50 of 6.1 ± 0.3 and

Emax of 39 ± 4% (P < 0.05, Table 1). A 2-way ANOVA also indicated a significant difference between the morphine CRC for MOPr-WT and MOPr-A6V (P < 0.0001, Figure

3). We have previously shown that the partial agonist buprenorphine has a significantly reduced efficacy for AC inhibition in cells expressing MOPr -N40D, a variant also located on the N-terminal region of MOPr (Knapman et al., In Press). Buprenorphine inhibited the

FSK response by 38 ± 7% in CHO-MOPr-WT cells, with pEC50 of 8.4 ± 0.3 (Table 1). In

CHO-MOPr-A6V cells, buprenorphine signalling was abolished (Figure 3). Maximum inhibition of the FSK response elicited by 10 μM pentazocine did not differ significantly between cells expressing MOPr-WT and MOPr-A6V (Table 1), however 2-way ANOVA showed a significant difference between the CRCs of CHO-MOPr-WT and CHO-MOPr-

A6V (P < 0.05, Figure 3). At concentrations higher than 10 μM, pentazocine has been reported to block K+ channels (Zhang & Cuevas, 2005), so concentrations above 10 μM were not tested despite pentazocine response not reaching a stable maximum at this concentration. For this reason, it was not possible to calculate pEC50 for pentazocine AC inhibition.

201

Figure 3: Morphine, buprenorphine and pentazocine inhibition of adenylyl cyclase and activation of ERK1/2 is compromised in CHO cells expressing MOPr-A6V. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) Morphine and pentazocine inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation was significantly decreased at MOPr-A6V compared with MOPr-WT (2-way ANOVA, P < 0.05). Buprenorphine signalling was abolished at MOPr-A6V. B) Morphine and pentazocine inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation was significantly decreased at MOPr-A6V compared with MOPr-WT (2-way ANOVA, P < 0.001). Buprenorphine signalling was abolished at MOPr-A6V. Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

202 A B

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT A6V A6V 60 60

40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Morphine] log M [Morphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT A6V A6V 60 60

40 40

20 20

0 0 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 [Buprenorphine] log M [Buprenorphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT A6V A6V 60 60

40 40

20 20

0 0 -8 -7 -6 -5 -8 -7 -6 -5 -4 [Pentazocine] log M [Pentazocine] Log M

203 We next tested the effect of MOPr-A6V on AC inhibition by the endogenous opioids β- endorphin, endomorphins-1 and 2, and met-enkephalin. β-endorphin signalling was negatively affected by the A6V variant (2-way ANOVA, P < 0.0001, Figure 4). In CHO-

MOPr-WT cells, β-endorphin inhibited the FSK response to a maximum of 73 ± 4%, with pEC50 of 6.8 ± 0.1. Maximum β-endorphin FSK inhibition was significantly decreased in

CHO-MOPr-A6V cells with Emax of 50 ± 4% (P < 0.05), and pEC50 of 6.4 ± 0.1 was also significantly different to MOPr-WT (Table 1). In cells expressing MOPr-A6V, CRCs for the endogenous opioid ligands endomorphin-1, endomorphin-2 and met-enkephalin differed significantly to CRCs of CHO-MOPr-WT (2-way ANOVA, P < 0.01, Figure 4).

Emax and pEC50 for endomorphin-1 in cells expressing MOPr-WT and MOPr-A6V were similar (Table 1). For endomorphin-2, Emax values for MOPr-WT and MOPr-A6V did not differ significantly, however pEC50 was affected at 8.2 ± 0.1 and 7.7 ± 0.1 respectively (P

< 0.05, Table 1). Likewise, maximum met-enkephalin inhibition of the FSK effect was similar between variants, but pEC50 values of 7.9 ± 0.1 at MOPr-WT and 7.2 ± 0.1 at

MOPr-A6V were significantly different (P < 0.05, Table 1).

CHO-MOPr cells were then challenged with the clinically important opioids fentanyl, methadone and oxycodone. The maximum inhibition of the FSK effect by fentanyl was similar in cells expressing MOPr-WT or MOPr-A6V, however 2-way ANOVA indicated a significant difference in the fentanyl CRCs (P < 0.01, Figure 5). There was also a significant effect of the A6V variant on the CRCs of methadone and oxycodone (2-way

ANOVA, P < 0.001, Figure 5). Maximum FSK inhibition by methadone was not different between MOPr-WT and MOPr-A6V, however pEC50s of 7.0 ± 0.1 and 6.4 ± 0.1 respectively were significantly different (P < 0.05, Table 1). Oxycodone Emax and pEC50 were not affected by MOPr-A6V (Table 1).

204 Opioid-mediated phosphorylation of ERK1/2

We next examined whether the A6V amino acid substitution has an impact on opioid signalling through another important pathway activated by MOPr, stimulation of ERK1/2 phosphorylation. We analysed the ability of opioids to stimulate ERK1/2 phosphorylation via MOPr-WT and MOPr-A6V in CHO cells, using a whole-cell ELISA assay. Opioid responses were normalised against 100 nM PMA applied for 10 minutes. The average

PMA response was similar in cells expressing MOPr-WT or MOPr-A6V, with corrected absorbance readings of 0.60 ± 0.06 and 0.53 ± 0.09, respectively (P > 0.05). Preincubation of cells with 1 µM naloxone for 5 mins blocked the response of EC80 concentrations of all opioids used in this assay. Application of DAMGO to cells treated overnight with

200ng/mL PTX, or in cells where receptor expression had not been induced did not elicit a response. DAMGO-stimulated ERK1/2 phosphorylation was significantly affected by the

A6V variant in CHO-MOPr cells when compared with WT expressing cells (2-way

ANOVA, P < 0.0001, Figure 2). DAMGO Emax for ERK1/2 phosphorylation was 70 ± 3% of the PMA response in CHO-MOPr-WT cells with a pEC50 of 8.6 ± 0.1, and in CHO-

MOPr-A6V cells, DAMGO Emax was 48 ± 2%, with pEC50 of 7.1 ± 0.4 (P < 0.05, Table 2).

ERK1/2 phosphorylation elicited by morphine was significantly altered at MOPr-A6V when compared with MOPr-WT (2-way ANOVA, P < 0.0001, Figure 3). Morphine activated ERK1/2 to a maximum of 58 ± 6% at MOPr-WT, and 39 ± 2% at MOPr-A6V (P

< 0.05). The pEC50 for morphine did not differ significantly between variants (Table 2).

Buprenorphine failed to elicit ERK1/2 phosphorylation in cells expressing MOPr-A6V

(Figure 3, Table 2). The CRC of ERK1/2 phosphorylation stimulated by pentazocine was altered at MOPr-A6V when compared with MOPr-WT (2-way ANOVA, P < 0.0001,

Figure 3), however Emax and pEC50 were not significantly different (Table 2).

205

Figure 4: Endogenous opioid inhibition of adenylyl cyclase and activation of ERK1/2 is significantly affected by the A6V variant in CHO cells expressing MOPr. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) β-endorphin, endomorphins 1 & 2, and met-enkephalin inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation was significantly different between MOPr-WT and MOPr-A6V (2-way ANOVA, P < 0.01). B) β-endorphin, endomorphins 1 & 2, and met-enkephalin activation of ERK1/2 was significantly different between MOPr- WT and MOPr-A6V (2-way ANOVA, P < 0.01). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

206 A B Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100

80 WT 80 WT A6V A6V 60 60

40 40

20 20

0 0 -8 -7 -6 -5 -9 -8 -7 -6 -5 [b-Endorphin] log M [b-Endorphin] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 120

100 80 WT WT A6V 80 A6V 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-1] log M [Endomorphin-1] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 120 100 80 WT WT A6V 80 A6V 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-2] log M [Endormorphin-2] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 120 100 80 WT WT A6V 80 A6V 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Met-Enkephalin] log M [Met-Enkephalin] log M

207 Table 2: Summary of opioid efficacy and potency in assays of ERK1/2 phosphorylation in CHO cells expressing MOPr-WT and MOPr-A6V.

pERK1/2 Emax (%) pEC50 Opioid WT A6V WT A6V Endomorphin-2 **** 106 ± 7 54 ± 7 8.2 ± 0.1 7.8 ± 0.2 Met-Enkephalin **** 103 ± 6 61 ± 7 8.1 ± 0.1 7.7 ± 0.2 Oxycodone **** 101 ± 6 57 ± 6 6.3 ± 0.1 6.1 ± 0.2 Fentanyl **** 98 ± 5 56 ± 4 8.2 ± 0.1 7.8 ± 0.1 Endomorphin-1 **** 97 ± 7 58 ± 5 8.2 ± 0.1 7.8 ± 0.1 Methadone **** 88 ± 6 46 ± 4 7.3 ± 0.1 6.7 ± 0.1 DAMGO **** 69 ± 5 38 ± 4 8.8 ± 0.2 7.7 ± 0.1 β-Endorphin **** 62 ± 2 56 ± 6 7.7 ± 0.1 6.4 ± 0.1 Morphine **** 58 ± 4 38 ± 3 7.2 ± 0.1 7.0 ± 0.1 Pentazocine *** 34 ± 9 12 ± 12 6.0 ± 0.4 6.0 ± 2.0 No No Buprenorphine 30 ± 4 9.1 ± 0.2 Response Response

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than endomorphin-2 are highlighted in red (1-way ANOVA, followed by Dunnett’s post-hoc test, corrected for multiple comparisons, P < 0.05). Opioids with concentration-response curves significantly different between variants are marked with * (2-way ANOVA, followed by Bonferroni post-hoc test, corrected for multiple comparisons. * P < 0.05, ** P < 0.01, *** P < 0.001,

**** P < 0.0001). MOPr-A6V Emax and pEC50 values significantly different from MOPr- WT are marked with * (unpaired Student’s T-test, P < 0.05).

208 The maximum ERK1/2 phosphorylation elicited by β-endorphin was similar between cells expressing MOPr-WT and MOPr-A6V however 2-way ANOVA showed a significant effect of MOPr-A6V on β-endorphin CRC (P < 0.0001, Figure 4). CRCs for endomorphin-1 and endomorphin-2 were also different at MOPr-A6V (2-way ANOVA, P < 0.0001, Figure 4), with Emax significantly decreased for both opioids, and endomorphin-1 pEC50 altered (P <

0.05, Table 2). Similarly, Emax for met-enkephalin stimulated ERK1/2 phosphorylation was decreased in CHO-MOPr-A6V cells compared with cells expressing MOPr-WT (P < 0.005,

Table 2), and CRCs were significantly different (2-way ANOVA, P < 0.0001, Figure 4).

The A6V variant had a negative impact on the ability of MOPr to stimulate ERK1/2 phosphorylation via fentanyl, methadone and oxycodone, with CRCs significantly different to

WT (2-way ANOVA, P < 0.0001, Figure 5). Maximal ERK1/2 phosphorylation by fentanyl, methadone and oxycodone was significantly decreased at MOPr-A6V (P < 0.05, Table 2).

The A6V variant had a negative effect on the signalling of most opioids tested for both AC inhibition and ERK1/2 phosphorylation, however the effect was more pronounced for

ERK1/2 phosphorylation. In assays of AC inhibition, the Emax and pEC50 of DAMGO, morphine, and β-endorphin were significantly affected by the A6V variant, and pEC50s of endomorphin-2, met-enkephalin and methadone were also affected (Table 1). In assays of

ERK1/2 phosphorylation, maximum opioid response was reduced by approximately 35 - 50% for every opioid tested with the exception of β-endorphin, where Emax was not affected but pEC50 was, and buprenorphine, where signalling was completely abolished (Table 2).

209

Figure 5: Fentanyl, methadone and oxycodone inhibition of adenylyl cyclase and activation of ERK1/2 is significantly affected by the A6V variant in CHO cells expressing MOPr. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. Adenylyl cyclase inhibition and levels of ERK phosphorylation were determined as described in the Methods. A) Fentanyl, methadone and oxycodone inhibition of forskolin-stimulated adenylyl cyclase hyperpolarisation was significantly different between MOPr-WT and MOPr-A6V (2-way ANOVA, P < 0.01). B) Fentanyl, methadone and oxycodone activation of ERK1/2 was significantly different between MOPr-WT and MOPr-A6V (2-way ANOVA, P < 0.0001). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

210

A B

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 120 100 80 WT WT A6V 80 A6V 60 60 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Fentanyl] log M [Fentanyl] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 80 WT WT A6V 80 A6V 60 60

40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Methadone] log M [Methadone] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 80 WT WT 80 A6V A6V 60 60

40 40

20 20

0 0 -8 -7 -6 -5 -4 -8 -7 -6 -5 -4 [Oxycodone] log M [Oxycodone] Log M

211 GIRK channel activation in AtT-20 cells

Activation of MOPr causes GIRK channel activation via Gβγ subunits, resulting in membrane hyperpolarisation. CHO-K1 cells do not express endogenous GIRK channels, so we assessed opioid mediated GIRK activation in AtT-20 cells stably transfected with hMOPr-WT and hMOPr-A6V as previously described (Knapman et al., 2013). In AtT-20 cells loaded with membrane potential sensitive dye, application of opioids resulted in a concentration-dependent decrease in fluorescence from baseline, corresponding to membrane hyperpolarisation from GIRK activation (Figure 6). DAMGO-stimulated GIRK activation in AtT20-MOPr-WT cells and AtT20-MOPr-A6V cells did not differ significantly (2-way ANOVA, Figure 7). DAMGO Emax was 34 ± 1% for cells expressing

MOPr-WT, and 31 ± 1% for cells expressing MOPr-A6V. DAMGO potency was also similar between variants, with pEC50 values of 8.4 ± 0.1 and 8.5 ± 0 (Table 3).

Morphine activated GIRK in AtT20-MOPr-WT cells to a maximum of 31 ± 1%, with pEC50 of 7.6 ± 0.1. In AtT20-MOPr-A6V cells, morphine Emax was significantly reduced, at 28 ± 1% (P < 0.05, Table 3). A 2-way ANOVA indicated a significant difference between MOPr-WT and MOPr-A6V morphine CRCs (P < 0.0001, Figure 7). Maximum buprenorphine-stimulated GIRK activation and pEC50 did not differ between cells expressing MOPr-WT and MOPr-A6V, however 2-way ANOVA showed a significant effect of the A6V variant on buprenorphine signalling overall (P < 0.05, Figure 7).

Interestingly, buprenorphine appears to be slightly more efficacious in AtT20-MOPr-A6V cells compared with AtT-MOPr-WT cells (Figure 7, Table 3). This is in direct contrast with the effect of the A6V variant on AC inhibition and ERK1/2 phosphorylation in CHO cells, where buprenorphine failed to elicit a response. In AtT20-MOPr-WT and AtT20-

MOPr-A6V cells, there was no difference in GIRK activation for pentazocine, β-endorphin or methadone (Figure 7, Table 3).

212 RFU (% baseline)

100

80

60 HBSS 40 WT 300 nM DAMGO

20 A6V 300 nM DAMGO

0 060 120 180 240 300 Time (secs)

Figure 6: DAMGO causes membrane hyperpolarisation in AtT-20 cells expressing MOPr-WT and MOPr-A6V. Raw trace showing decrease in fluorescent signal following application of 300 nM DAMGO or 3 µM buprenorphine at 120 sec, corresponding to membrane hyperpolarisation from GIRK activation.

213

Figure 7: Morphine activates GIRK less effectively in AtT20 cells expressing MOPr- A6V. GIRK activation was determined as described in the Methods. Morphine and buprenorphine activation of GIRK channels was significantly different between MOPr-WT and MOPr-A6V (2-way ANOVA, P < 0.05). Morphine Emax was decreased from 31 ± 1% in AtT20-MOPr-WT to 28 ± 1% in AtT20-MOPr-A6V (Student’s T-test, P < 0.05).

Buprenorphine Emax and pEC50 did not differ significantly between MOPr-WT and MOPr- A6V. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

214

D Fluorescence (%) D Fluorescence (%) 50 50

WT WT 40 40 A6V A6V 30 30

20 20

10 10

0 0 -9 -8 -7 -6 -9 -8 -7 -6 -5 [DAMGO] log M [Morphine] log M

D Fluorescence (%) D Fluorescence (%) 50 50

WT WT 40 40 A6V A6V 30 30

20 20

10 10

0 0 -8 -7 -6 -5 -8 -7 -6 -5 [Buprenorphine] log M [Pentazocine] log M

D Fluorescence (%) D Fluorescence (%) 50 50

WT WT 40 40 A6V A6V 30 30

20 20

10 10

0 0 -8 -7 -6 -8 -7 -6 -5 [b-Endorphin] log M [Methadone] log M

215

Table 3: Summary of opioid efficacy and potency in assays of GIRK activation in AtT-20 cells expressing MOPr-WT and MOPr-A6V.

GIRK activation Emax (%) pEC50 Opioid WT A6V WT A6V DAMGO 34 ± 1 31 ± 1 8.4 ± 0.1 8.5 ±0.1 β-Endorphin 35 ± 2 32 ± 1 7.0 ± 0.1 7.0 ± 0.1 Methadone 33 ± 1 30 ± 1 7.3 ± 0.1 7.4 ± 0.1 Morphine **** 31 ± 1 28 ± 1 7.6 ± 0.1 7.6 ± 0.1 Buprenorphine * 21 ± 2 22 ± 1 7.0 ± 0.1 7.3 ± 0.1 Pentazocine 8 ± 1 7 ± 1 7.0 ± 0.1 7.4 ± 0.1

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT. Opioids with Emax significantly lower than DAMGO are highlighted in red (1-way ANOVA, Dunnett’s post-hoc test, corrected for multiple comparisons, P < 0.05). Opioids with concentration-response curves significantly different between variants are marked with * (2-way ANOVA, followed by Bonferroni post-hoc test, corrected for multiple comparisons. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001). MOPr-A6V Emax and pEC50 values significantly different from MOPr-WT are marked with * (unpaired Student’s T-test, P < 0.05).

216 DISCUSSION and CONCLUSIONS

We have shown that a relatively commonly occurring MOPr variant, A6V has a significant, detrimental impact on the ability of MOPr to couple to effector pathways in

CHO cells. In assays of AC inhibition, buprenorphine signalling was abolished at the A6V variant. Most opioids tested had reduced potency in CHO-MOPr-A6V cells compared with

CHO-MOPr-WT cells, and the response at maximal concentrations of opioid was significantly reduced for morphine, β-endorphin, methadone and oxycodone. In assays of

ERK1/2 phosphorylation, buprenorphine also failed to elicit a response in cells expressing

MOPr-A6V. Maximum ERK1/2 phosphorylation was significantly decreased for DAMGO and all other opioids tested in CHO-MOPr-A6V cells, with the exception of β-endorphin.

A significant effect of A6V was also apparent at sub-maximal concentrations of β- endorphin, as well as DAMGO, endomorphin-1, endomorphin-2, met-enkephalin, fentanyl and methadone. In contrast to assays of AC inhibition and ERK1/2 phosphorylation, MOPr coupling to GIRK activation did not appear to be significantly affected by the A6V variant.

In AtT20-MOPr-A6V cells, GIRK activation was similar for all opioids tested, with the exception of morphine for which a significant effect of the variant was observed at submaximal concentrations, although it seems ulikely this would translate to a significant physiological effect. Most strikingly, buprenorphine-stimulated GIRK activation was not affected by the A6V variant when expressed in AtT-20 cells.

In the absence of spare receptors, even modest differences in levels of receptor expression could have a significant impact on the signalling profile of MOPr variants. There is no information on the effects of the A6V variant on receptor expression in humans or animal models. However, in our assays, MOPr expression was similar for both variants for CHO-

K1 cells. Radioligand binding assays showed that CHO-MOPr-A6V cells had similar surface receptor expression as CHO-MOPr-WT cells, and thus receptor expression levels

217 are unlikely to have contributed to a decrease in opioid signalling at MOPr-A6V. For

AtT20-MOPr cells, receptor expression was not directly measured, however the use of the

FlpIn system for receptor transfection ensures that the receptor construct is inserted into the same location in the genome. All cells are thus subjected to a similar transcriptional environment (Sauer, 1994). The efficacy of the partial agonists buprenorphine and pentazocine was similar in AtT20 cells expressing MOPr-WT and MOPr-A6V, suggesting similar expression levels.

Our findings suggest that the alanine to valine amino acid change on the N-terminus of

MOPr may affect the ability of opioid ligands to bind to MOPr and effectively transduce signal to intracellular effectors. Genetic variation of GPCRs may affect receptor function by altering GPCR conformation, with conformational changes potentially affecting the ability of ligands to bind to the receptor, and/or affecting the efficacy of the resulting ligand/receptor complex in coupling to associated effector molecules (Kenakin & Miller,

2010; Pineyro et al., 2007). Most studies examining ligand interaction with GPCRs have focused on regions that form the ligand-binding pocket, namely the transmembrane helices and extracellular loops, and the recently published crystal structures of MOPr, DOPr and

KOPr do not include the N-terminus (Granier et al., 2012; Manglik et al., 2012; Wu et al.,

2012). The A6V residue change is located at the distal end of the N-terminal region, and thus it is somewhat surprising that this variant has such a striking impact on MOPr function. However, several studies investigating the functional consequences of the common MOPr variant N40D, which is also present in the N-terminal region of MOPr, have reported altered N40D signalling (reviewed in Knapman & Connor, 2014). We have recently shown that in CHO cells, the MOPr-N40D variant selectively decreases buprenorphine efficacy for AC inhibition and ERK1/2 phosphorylation. Furthermore, buprenorphine potency for GIRK activation was decreased in AtT20-MOPr-N40 cells

218 (Knapman et al.,2014). In the present study, buprenorphine was also the most markedly affected opioid, with its activity completely abolished in CHO-MOPr-A6V cells for both

AC inhibition and ERK1/2 phosphorylation. The N40D variant of MOPr removes a putative glycosylation site, which has been associated with decreased levels of mRNA and receptor expression in animal models and post-mortem human brain (Oertel et al., 2009;

Zhang et al., 2005). The A6V variant does not affect a MOPr glycosylation site, and receptor expression was equivalent between cell lines, suggesting that genetic changes in the N-terminal region of MOPr may be capable of affecting receptor function independently of receptor glycosylation.

Importantly, the N-terminal region has been shown to undergo activation-dependent changes in MOPr and other GPCRs. Antibodies generated against the N-terminal domain of MOPr show differential recognition of inactive and activated receptors, suggesting this region undergoes conformational changes upon ligand-binding (Gupta et al., 2007, Gupta et al., 2008). Changes in the C-terminal region of MOPr also affected N-terminal antibody binding, implying that all domains of MOPr may be affected by conformational changes associated with receptor coupling to associated G-proteins. A similar effect was seen in other GPCRs examined in this study including the δ-opioid receptor and the cannabinoid

CB1 receptor (Gupta et al., 2007). Signalling in some other GPCRs is also affected by N- terminal SNPs. Constitutive and agonist-stimulated activity of the 5-HT2B receptor was increased by an R > G aa substitution at an equivalent position on the N-terminal tail to

MOPr-A6V (Belmer et al., 2014). It is perhaps not surprising that the residue change from a large, charged arginine molecule to a small neutral glycine molecule has a significant impact on receptor signalling. In the present study, the A6V substitution on MOPr is a relatively small change with a single methylene addition, and given its position, would generally not be expected to have a significant effect on receptor function. Nonetheless, the

219 dramatic impact of this alteration on MOPr signalling underscores the potential importance of this region to global GPCR conformation and function (Belmer et al. 2014). N-terminal polymorphisms in the melanocortin-4 receptor have also been shown to increase constitutive activity (Srinivasan et al., 2004). The results of these studies suggest the N- terminal region may undergo substantial structural changes upon receptor activation, and that structural changes arising from genetic variation have the ability to significantly affect receptor function.

Intriguingly, although opioid signalling was markedly decreased in CHO cells expressing

MOPr-A6V, in AtT20-MOPr cells the A6V variant had little effect on the ability of opioids to activate GIRK channels. This was particularly apparent for buprenorphine, which failed to inhibit AC or stimulate ERK1/2 in CHO cells expressing MOPr-A6V, yet activated GIRK channels to a similar extent as MOPr-WT in AtT20 cells. This data suggests that the A6V polymorphism may result in pathway selective losses of function.

MOPr inhibition of AC is mediated by Gαi/o subunits of the G-protein heterotrimer, whereas GIRK activation occurs via Gβγ subunits (Law et al., 2000). ERK1/2 phosphorylation can occur via multiple pathways, and may involve Gα and Gβγ-coupled processes as well as G-protein independent pathways such as β-arrestin (Luttrell et al.,

2005). At present it is not known which specific subtypes of Gi/o protein couple MOPr to inhibition of AC or activation of ERK in CHO cells, or activation of GIRK in AtT-20 cells.

Thus, it is not clear if the selective loss of efficacy in signal transduction mediated by A6V in CHO cells is because the MOPr-A6V couples less well to G-mediated process than

G-mediated signalling, or whether it is disruption of coupling to a specific subtype of

Gi/o that is responsible for AC inhibition and ERK phosphorylation in CHO cells, but which is not involved in GIRK activation in AtT-20 cells. The apparently greater detrimental effect of the A6V variant on ERK1/2 phosphorylation compared with AC

220 inhibition in CHO-MOPr cells may also be due to differences in the effector molecules required for activation of these pathways. The apparently greater loss of efficacy for some ligands such as buprenorphine at MOPr-A6V signalling to AC and ERK may also reflect some functional selectivity towards GIRK activation over AC inhibition or ERK1/2 phosphorylation, although studies of AC inhibition and ERK1/2 phosphorylation in AtT-20 cells are necessary to definitively establish the presence of any functional selectivity. Each of the assays utilized here is relatively rapid, with a maximum 5 minute time-course, and a degree of opioid receptor desensitisation can occur over this time. Conceivably, a reduction in AC inhibition or pERK levels at 5 minutes could represent enhanced receptor desensitisation. However, we have not seen any evidence for enhanced desensitisation of

MOPr-A6V activation of GIRK at 5 minutes (Santiago and Connor, unpublished observations), and no differences in morphine or DAMGO-induced internalisation were reported for A6V expressed on the MOR1A background (Ravindranathan et al., 2009), suggesting that agonist-induced regulation of MOPr-A6V is not obviously accelerated compared with MOPr-WT.

There are no functional studies investigating the consequences of the A6V variant on acute opioid signalling published. The results from our study demonstrate a significant impact of the A6V variant on the ability of MOPr to activate associated signalling pathways and warrant further investigation with functional studies examining other MOPr signalling pathways using alternative cellular backgrounds.

The A6V variant is present at allelic frequencies of more than 20% in some populations, thus any differences in signalling arising from the A6V variant could be of clinical significance. The A6V variant markedly decreased the effect of a number of clinically prescribed opioids including morphine, one of the most commonly prescribed strong

221 analgesics worldwide (Hanks et al., 2001; Pergolizzi et al., 2008; Zernikow et al., 2009).

Furthermore, the signalling of the endogenous opioids β-endorphin, endomorphins 1 and 2, and met-enkephalin was significantly compromised at MOPr-A6V. Some studies have suggested a higher frequency of the A6V variant in substance abusing populations, and disruption of the normal function of the endogenous opioid system caused by the A6V variant could conceivably contribute to a higher predisposition to substance abuse

(Berrettini et al., 1997; Compton et al., 2003; Crowley et al., 2003; Crystal et al., 2010;

Rommelspacher et al., 2001).

Our results demonstrate that the A6V variant has the potential to affect MOPr function across cell types and signalling pathways. Most carriers of the 17T allele will be heterozygous with 17C, and at present it is not known how co-expressed MOPr-WT and

MOPr-A6V receptors will signal. There is still considerable uncertainty about whether

MOPr and other Class A GPCR form obligate dimers in native tissue (Malik et al., 2013;

Herrick-Davis 2013), so whether cells expressing A6 and V6 receptors will have a mixture of homo- and heterodimers, or 2 populations of receptor which signal independently is unknown. To date, the effect of MOPr variants expressed together with MOPr-WT has been largely unexplored, largely owing to the technical difficulty of co-expressing accurately quantified and equivalent amounts of each variant. Ravindranathan et al.

(2009) co-expressed HA-tagged MOPr-WT and FLAG-tagged MOPr variants in HEK293 cells and reported the internalisation of both MOPr-WT and MOPr-L85I in response to morphine, suggesting the formation of functional dimers with L85I showing a dominant phenotype. On the other hand, when MOPr-WT and MOPr-R181C were co-expressed,

MOPr-WT internalised independently in response to DAMGO, indicating that this variant either fails to form dimers or forms unstable dimers with MOPr-WT (Ravindranathan et al., 2009). The effect of co-expression on receptor expression and signalling was not

222 measured, and the physiological relevance is unknown. In any case, potential interactions between WT- and A6V-MOPr require further examination, and the investigation of opioid response in A6V carriers in a clinical setting would also be of great interest.

The A6V variant is highly prevalent in some populations, occurring at allelic frequencies of up to 20% (Crowley et al., 2003; Kapur et al., 2007; Rommelspacher et al., 2001; Tan et al., 2003). A decrease in the efficacy of opioid analgesics in A6V carriers could result in a significant proportion of the population receiving inadequate or inappropriate analgesic therapy. Understanding how the A6V variant affects opioid signalling is an important element in predicting individual response to opioids, and the potential clinical or phenotypic consequences of altered MOPr function. Although individual opioid response is influenced by a number of genetic and epigenetic factors, knowledge of the effect of individual MOPr genotype on opioid response could provide valuable insight into the most effective form of analgesic therapy, particularly in the case of A6V carriers, where opioid efficacy may be significantly diminished. Understanding how potential disruptions in endogenous opioid signalling affect individual phenotype could also be beneficial in elucidating the role of endogenous opioids in physiological processes such as nociception, reward and addiction.

Acknowledgements:

This study was supported the National Health and Medical Research Council of Australia project grant 1011979 to MC. AK and MS were recipients of Postgraduate Scholarships from Macquarie University, topped up by the Australian School of Advanced Medicine.

We thank Mac Christie and Yan Ping Du for their assistance with the binding assays, and

Courtney Breen for providing a protocol and advice on the ERK ELISA.

223 Author Contributions

AK, MS and MC designed and analysed experiments, AK and MS conducted the experiments. AK and MC conceived the study and wrote the paper, all authors have seen the final manuscript.

224 REFERENCES

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al.

(2013). The concise guide to PHARMACOLOGY 2013/14: G protein-coupled receptors.

Br J Pharmacol 170: 1459-1581.

Belmer A, Doly S, Setola V, Banas SM, Moutkine I, Boutourlinsky K et al. (2014). Role of the N-terminal region in G protein-coupled receptor functions: Negative modulation revealed by 5-HT2B receptor polymorphisms. Mol Pharmacol 85: 127-138.

Berrettini WH, Hoehe MR, Ferraro TN, DeMaria PA, Gottheil E (1997). Human mu- opioid receptor gene polymorphisms and vulnerability to substance abuse. Addiction Biol

2: 303-308.

Boom M, Niesters M, Sarton E, Aarts L, Smith TW, Dahan A (2012). Non-analgesic effects of opioids: opioid-induced respiratory depression. Curr Pharm Des 18: 5994-6004.

Compton P, Geschwind DH, Alarcon M (2003). Association between human mu-opioid receptor gene polymorphism, pain tolerance, and opioid addiction. Am J Med Genet B

Neuropsychiatr Genet 121B: 76-82.

Cox BM, Christie MJ, Devi L, Toll L, Traynor JR (2014). Challenges for opioid receptor nomenclature. Br J Pharmacol, In Press.

Crist RC, Berrettini WH (2013). Pharmacogenetics of OPRM1. Pharmacol Biochem

Behav, In Press doi: 10.1016/j.pbb.2013.10.018.

225 Crowley JJ, Oslin DW, Patkar AA, Gottheil E, DeMaria PA Jr, O’Brien CP et al. (2003).

A genetic association study of the mu opioid receptor and severe opioid dependence.

Psychiatr Genet 13: 169-73.

Crystal HA, Hamon S, Randesi M, Cook J, Anastos K, Lazar J (2012). A C17T polymorphism in the mu opiate receptor is associated with quantitative measures of drug use in African American women. Addict Biol 17: 181-91.

Diatchenko L, Robinson JE, Maixner W (2011). Elucidation of mu-opioid gene structure:

How genetics can help predict therapeutic response to opioids. Eur J Pain Suppl 5: 433-48.

Fortin JP, Ci L, Schroeder J, Goldstein C, Montefusco MC, Peter I et al. (2010). The μ- opioid receptor variant N190K is unresponsive to peptide agonists yet can be rescued by small-molecule drugs. Mol Pharmacol 78: 837-845.

Granier S, Manglik A, Kruse AC, Kobilka TS, Thian FS, Weis WI et al. (2012). Structure of the δ-opioid receptor bound to . Nature 485: 400-404.

Gupta A, Decaillot FM, Gomes I, Tkalych O, Heimann AS, Ferro ES et al. (2007).

Conformation state-sensitive antibodies to G-protein-coupled receptors. J Biol Chem 282:

5116-5124.

Gupta A, Rozenfeld R, Gomes I, Raehal KM, Decaillot FM, Bohn LM et al. (2008). Post- activation-mediated changes in opioid receptors detected by N-terminal antibodies. J Biol

Chem 283: 10735-10744.

226 Kapur S, Sharad S, Singh RA, Gupta AK (2007). A118G polymorphism in mu opioid receptor gene (oprm1): association with opiate addiction in subjects of Indian origin. J

Integr Neurosci 6: 511-522.

Kenakin T, Miller LJ (2010). Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery.

Pharmacol Rev 62: 265-304.

Kim K-J, Kim H-E, Lee K-W, Han W, Yi, M-J, Jeong J et al. (2004). Two-promoter vector is highly efficicent for overproduction of protein complexes. Protein Sci 13: 1698-1703.

Klepstad P, Fladvad T, Skorpen F, Bjordal K, Caraceni A, Dale O et al. (2011). Influence from genetic variability on opioid use for cancer pain: A European genetic association study of 2294 cancer pain patients. Pain 152: 1139-1145.

Knapman A, Abogadie F, McIntyre P, Connor M (2013). A real-time, fluorescence-based assay for measuring μ-opioid receptor modulation of adenylyl cyclase activity in Chinese hamster ovary cells. J Biomol Screen 19: 223-231.

Knapman A, Connor M (2014). Cellular signalling of non-synonymous single nucleotide polymorphisms of the human µ-Opioid receptor (OPRM1). Br J Pharmacol, In Press doi:

10.1111/bph.12644

Knapman A, Santiago M, Connor M (2014). Buprenorphine signalling is compromised at the N40D polymorphism of the human µ-opioid receptor in vitro. Br J Pharmacol (In

Press).

227 Knapman A, Santiago M, Du YP, Bennallack PR, Christie MJ, Connor M (2013). A continuous, fluorescence-based assay of μ-opioid receptor activation in AtT-20 cells. J

Biomol Screen 18: 269-76.

Law PY, Wong YH, Loh HH (2000). Molecular mechanisms and regulation of opioid receptor signalling. Annu Rev Pharmacol Toxicol 40: 389-430.

Lotsch J, Geisslinger G (2005). Are μ-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol Med 11: 82-89.

Luttrell LM (2005). Composition and function of g protein-coupled receptor signalsomes controlling mitogen-activated protein kinase activity. J Mol Neurosci 26L 253-264.

Mague SD, Blendy JA (2010). OPRM1 SNP (A118G): involvement in disease development, treatment response, and animal models. Drug Alcohol Depend 108: 172-182.

Malik RU, Ritt M, DeVree BT, Neubig RR, Sunahara RK, Sivaramakrishnan S (2013).

Detection of G protein-selective G protein-coupled receptor (GPCR) conformations in live cells. J Biol Chem 288: 17167-17178.

Manglik A, Kruse AC, Kobilka TS, Thian FS, Mathiesen JM, Sunahara RK et al. (2012).

Crystal structure of the μ-opioid receptor bound to a morphinan antagonist. Nature 485:

321-326.

228 Matthes HWD, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I et al. (1996).

Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the µ-opioid-receptor gene. Nature 383: 819-823.

Mura E, Govoni S, Racchi M, Carossa V, Ranzani GN, Allegri M, van Schaik RHN

(2013). Consequences of the 118A>G polymorphism in the OPRM1 gene: translation from bench to bedside? J Pain Res 6: 331-53.

Nozaki C, Kamei J (2007). Involvement of mu1-opioid receptor on oxycodone-induced antinociception in diabetic mice. Eur J Pharmacol 560: 160-162.

Oertel BG, Kettner M, Scholich K, Renne C, Roskam B, Geisslinger G et al. (2009). A common human μ-opioid receptor genetic variant diminishes the receptor signalling efficacy in brain regions processing the sensory information of pain. J Biol Chem 284:

6530-6535.

Pfutzner W (2008). Retroviral bicistronic vectors. Drug News Perspect 21: 473-580.

Pineyro G, Archer-Lahlou E (2007). Ligand-specific receptor states: implications for opiate receptor signalling and regulation. Cell Signal 19: 8-19.

Ravindranathan A, Joslyn G, Robertson M, Schuckit MA, Whistler JL, White RL (2009).

Functional characterisation of human variants of the mu-opioid receptor gene. Proc Natl

Acad Sci USA 106: 10811-10816.

229 Rommelspacher H, Smolka M, Schmidt LG, Samochowiec J, Hoehe MR (2001). Genetic analysis of the mu-opioid receptor in alcohol-dependent individuals. Alcohol 24: 129-135.

Sauer B (1994). Site-specific recombination: Developments and applications. Curr Opin

Biotechnol 5: 521-527.

Skorpen F, Laugsend EA (2008). Variable response to opioid treatment: any genetic predictors in sight? Palliat Med 22: 310-327.

Somogyi AA, Barrat DT, Coller JK (2007) Pharmacogenetics of opioids. Clin Pharmacol

Ther 81: 429-444.

Srinivasan S, Lubrano-Berthelier C, Govaerts C, Picard F, Santiago P, Conklin BR et al

(2004). Constitutive activity of the melanocortin-4 receptor is maintained by its N-terminal domain and plays a role in energy homeostasis in humans. J Clin Invest 114: 1158-1164.

Tan EC, Tan CH, Karupathivan U, Yap EP (2003). Mu opioid receptor gene polymorphisms and heroin dependence in Asian populations. Neuroreport 14: 569-572.

Virk MS, Arttamangkul S, Birdsong WT, Williams JT (2009). Buprenorphine is a weak partial agonist that inhibits opioid receptor desensitisation. J Neurosci 29: 7341-7348.

Walter C, Lotsch J (2009). Meta-analysis of the relevance of the OPRM1 118A>G genetic variant for pain treatment. Pain 146: 270-275.

230 Wu H, Wacker D, Mileni M, Katritch V, Han GW, Vardy E et al. (2012). Structure of the human κ-opioid receptor in complex with JDTic. Nature 485: 327-332.

Zhang H, Cuevas J (2005). σ-receptor activation blocks potassium channels and depresses neuroexcitability in rat intracardiac neurons. J Pharmacol Exp Ther 313: 1387-1396.

Zhang Y, Wang D, Johnson AD, Papp AC, Sadee W (2005). Allelic expression imbalance of human mu opioid receptor (OPRM1) caused by variant A118G. J Biol Chem 280:

32618-24.

231

232 Chapter 7

Mu-opioid receptor polymorphisms differentially affect receptor signalling via adenylyl cyclase inhibition and ERK phosphorylation

Alisa Knapman and Mark Connor

This chapter has been prepared and formatted according to research article guidelines for the British Journal of Pharmacology. Mark Connor assisted with experimental design and data analysis. All other work is my own.

233 INTRODUCTION

Single nucleotide polymorphisms (SNPs) in genes coding for G protein-coupled receptors

(GPCRs) can result in aberrant signalling profiles of these receptors (Thompson et al.,

2008; Vassart & Costagliola, 2011). An amino acid substitution arising from a non- synonymous SNP may alter the structure and/or conformation of a GPCR in such a way as to affect the ability of a ligand to bind to the receptor and elicit a response. The μ-opioid receptor (MOPr) is a GPCR that is the primary target for most clinically prescribed opioid analgesics including morphine, oxycodone and fentanyl (Matthes et al., 1996; Pawson et al., 2014). Genotyping of the human MOPr gene (OPRM1) has revealed the presence of a number of non-synonymous SNPs within the population (Lotsch & Geisslinger, 2005).

Individual response to opioids is highly variable, and as such naturally occurring MOPr variants are possible contributors to interindividual differences in opioid response

Differences in individual opioid response have been associated with the two most common

OPRM1 SNPs, the N40D variant and the A6V variant, both of which occur in the N- terminal region of MOPr (Lotsch & Geisslinger, 2005; Somogyi et al., 2007). Other rare

OPRM1 SNPs, present in less than 1% of the population, have received far less attention.

These include the L85I variant, present in the first transmembrane domain (TM1) of

MOPr, the R181C variant on the second intracellular loop (ICL2), and the R260H, R265H and S268P variants, which affect the third intracellular loop (ICL3) of the receptor

((Lotsch & Geisslinger, 2005; Ravindranathan et al., 2009). The clinical effects of these variants in humans are not known due to their rarity within the population, yet the potential insights that MOPr SNPs can provide into receptor function has prompted several studies investigating the consequences of rare OPRM1 SNPs in vitro.

The potential for MOPr SNPs to contribute to the clinical variability of opioid responses is very real, and has been demonstrated for several GPCRs (Thompson et al., 2008; Zhang et

234 al., 2013a). Like all GPCRs, the MOPr has many active conformations, which may differentially activate various downstream effector and regulatory molecules (Pinyero et al., 2007; Kenakin & Miller, 2010; Manglik et al., 2012). Single amino acid substitutions can lead to subtle or profound changes in receptor conformation, potentially affecting the signalling pathways activated (Abrol et al., 2013; Cox, 2013). Ligands may also exhibit specific patterns of signalling and receptor regulation by preferentially binding to and stabilising subsets of MOPr conformational states (Kenakin, 2002; Massotte et al., 2002;

Saidak et al., 2006). SNPs resulting in single amino acid changes therefore have the potential to affect MOPr signalling globally or in a ligand-dependent manner by affecting the ability of a ligand to bind to the receptor, altering the conformation of the ligand- receptor complex and/or affecting the ability of this complex to couple to G-proteins and associated signalling or regulatory pathways of MOPr.

In this study, we investigated the effect of the L85I, R181C, R260H, R265H and S268P variants on the ability of human MOPr to inhibit adenylyl cyclase and stimulate pERK phosphorylation in CHO cells, using 11 clinically important and/or structurally distinct opioid ligands. We found significant enhancement of adenylyl cyclase inhibition by selected opioids at MOPr-L85I, but this enhancement was not evident in assays of ERK1/2 phosphorylation. MOPr-R260H had a markedly negative impact on opioid signalling in assays of AC inhibition and ERK1/2 phosphorylation, whereas MOPr-R265H signalling was similar to MOPr-WT for most opioids tested.

235 METHODS

MOR transfection and cell culture

CHO-FRT-TRex cells were stably transfected with a pcDNA5 construct encoding the haemagglutinin (HA)-tagged human -opioid receptor cDNA together with the pOG44

(Flp recombinase plasmid) using the transfectant Fugene (Promega), as described previously (Knapman et al., 2014). The HA-tagged human wild-type -opioid receptor

(MOPr-WT) and the six variant HA-tagged -opioid receptors were synthesised by

Genscript (Piscataway, New Jersey, USA). The MOPr variants examined were the A6V variant (MOPr-A6V), the L85I variant (MOPr-L85I), the R181C variant (MOPr-R181C), the R260H variant (MOPr-R260H), the R265H variant (MOPr-265H) and the S268P variant (MOPr-S268P). Cells expressing MOPr were selected using hygromycin B (500

µg/mL). Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) containing

10% FBS, 100U penicillin/streptomycin and 500 g/mL hygromycin B up to passage 5.

Hygromycin concentration was reduced to 200 g/mL beyond passage 5.

CHO-MOPr cells were passaged at 80% confluency as required. Assays were carried out on cells up to 30 passages. Cells for assays were grown in 75 cm2 flasks and used at greater than 90% confluence.

MOPr receptor expression

Surface expression of all MOPr variants was determined on intact CHO-MOPr cells by incubation with [3H]DAMGO (0.125–16 nM; PerkinElmer, Waltham, MA) at 4 °C in 50 mM Tris-Cl (pH 7.4) for 2 h. Briefly, approximately 1 × 105 cells were plated in a 24-well plate and grown overnight. Cells were then washed gently twice with 50 mM Tris-Cl (pH

7.4) and placed on ice, and incubated for 2h with increasing concentrations of

236 [3H]DAMGO. Nonspecific binding was determined in the presence of unlabelled DAMGO

(10 μM). At the end of the incubation, plated cells were washed three times with 50 mM

Tris-Cl (pH 7.4) at 4 °C. Cells in each well were then digested for 1 h at room temperature with 100 μL of 1N NaOH. 100 μL 1N HCl was added to each well and collected into scintillation vials and bound ligand determined using a liquid scintillation counter (Packard

Tricarb, Perkin Elmer, Waltham MA, USA). Specific binding was plotted, and Kd and Bmax for each clone were determined using GraphPad Prism (GraphPad Software, La Jolla, CA).

Protein concentration was determined with a BCA Protein Assay Kit (Pierce) according to the manufacturer’s instructions.

Membrane potential assays

The day before the assay, cells were detached from the flask with trypsin/EDTA (Sigma) and resuspended in 10 ml of Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin and 15 mM glucose. MOPr receptor expression was induced in

CHO-MOPr cells with 2 g/mL tetracycline 24 hrs prior to the assay. The cells were plated in a volume of 90 µl in black walled, clear-bottomed 96-well microplates (Corning), and

o incubated overnight at 37 C in ambient CO2. Membrane potential was measured using a

FLIPR Membrane Potential Assay kit (blue) from Molecular Devices. The dye was reconstituted with assay buffer containing (in mM), NaCl 145, HEPES 22, Na2HPO4

0.338, NaHCO3 4.17, KH2PO4 0.441, MgSO4 0.407, MgCl2 0.493, CaCl2 1.26, glucose

5.56 (pH 7.4, osmolarity 315 ± 5). Prior to the assay, cells were loaded with 90 l/well of the dye solution without removal of the L-15, giving an initial assay volume of 180

l/well. Plates were then incubated at 37°C in ambient CO2 for 60 minutes. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at 565 nm. Baseline readings were taken every 2 seconds for at least 2 minutes, at which time either drug or vehicle was

237 added in a volume of 20 μL. Further additions were made in volumes of 20 µl, as indicated. The background fluorescence of cells without dye or dye without cells was negligible. Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle addition. When used, the final concentration of the solvents DMSO or EtOH was not more than 0.1%, and these concentrations did not produce a signal in the assay.

ELISA of ERK1/2 phosphorylation

Opioid-mediated ERK1/2 phosphorylation was measured by ELISA, as previously described (Knapman et al., In Press). Briefly, 24 hrs before the assay, CHO-MOPr cells were plated in 96-well clear microplates (Falcon) and receptor expression was induced as for the AC inhibition assay. Cells were serum-starved for 1 hr in 40 μL serum-free L-15 supplemented with 5% bovine serum albumin (BSA) before the assay. Cells were treated with drug or vehicle diluted in serum-free L-15 (40 µl added) with no BSA added to minimise background. Preliminary time-course experiments indicated that a 5 min drug treatment was optimal to produce robust ERK1/2 phosphorylation without inducing desensitisation, thus all drug treatments were for 5 min. After drug application, the reaction was stopped by inverting the plates to remove the drug solution, placing the plates on ice, and immediately fixing the cells with 4% paraformaldehyde for 15 min at RT. Cells were washed 3 times with 300μL PBS, then permeabilized with 0.1% Triton-X in PBS for 30 min at RT. Triton-X was removed, and cells were incubated for 2 hr at RT with blocking solution consisting of 5% BSA in PBS with 0.01% Tween-20 (PBS-T). Blocking solution was removed, then cells were incubated overnight at 4°C with a 1:500 dilution of rabbit α- phospho-p44/42 MAPK (Thr202/Tyr204) antibody in PBS-T with 1% BSA. Cells were washed 3 times with 300μL PBS-T, and incubated with 1:5000 α-rabbit IgG HRP-linked antibody in PBS-T with 1% BSA for 2 hr. Cells were washed 4 times with 300 μL PBS-T,

238 and incubated with 3,3’,5,5’-tetramethylbenzidine (TMB; Sigma) at RT in the dark for 45 min. The reaction was stopped with 1M HCl. Absorbance was read at 450 nm using a

BMG Pherastar FS microplate reader. Cells were then stained with 0.5μg/mL DAPI for 10 min at RT, and washed 3 times with 300 μL PBS-T. Fluorescence was read in the Pherastar microplate reader with cells excited at a wavelength of 358 nm and emission measured at

461 nm. Absorbance readings were normalised to DAPI staining to account for any differences in cell density between wells. Readings were then normalised to the response of cells treated with 100nM PMA for 10 min.

Drugs and Chemicals

Tissue culture reagents and buffer salts were from Life Technologies or Sigma unless otherwise noted. Tyr-D-Ala-Gly-N-MePhe-Gly-ol acetate (DAMGO), endomorphin-1, endomorphin-2 and met-enkephalin were purchased from Auspep (Tullamarine, Australia).

Morphine, fentanyl and pentazocine were a kind gift from the Department of

Pharmacology, University of Sydney. Buprenorphine and oxycodone were from the

National Measurement Institute (Lindfield, Australia). β-endorphin was from Genscript

(Piscataway, New Jersey, USA). Forskolin and naloxone were from Ascent

Pharmaceuticals (Bristol, UK). 1,9-dideoxyforskolin and tetraethylammonium (TEA) were from Sigma Aldrich (Castle Hill, Australia). Pertussis toxin (PTX) was from Tocris

Bioscience (Bristol, UK). Phospho-ERK1/2 antibody (Catalog #9101) and anti-rabbit IgG

HRP-lined antibody (Catalog #7074) were from Cell Signalling Technologies, Danvers,

Massachusetts, USA.

Data

Unless otherwise noted, data is expressed as mean ± s.e.m. of at least 5 determinations made in duplicate or triplicate. Concentration response curves were fit with a 4 parameter logistic equation using Graphpad Prism (Graphpad). Statistical comparisons were made

239 with one-way ANOVA followed by Dunnett’s post-hoc test corrected for multiple comparisons, or an unpaired Student’s T-test not corrected for multiple comparisons. P <

0.05 was considered significant. All channel and receptor nomenclature is consistent with the British Journal of Pharmacology Guide to Receptors and Channels (Alexander et al.,

2013).

RESULTS

MOPr Expression in CHO-K1 cells

CHO-K1 cells were stably transfected with MOPr-WT or with MOPr-L85I, MOPr-R181C,

MOPr-R260H, MOPr-R265H or MOPr-S268P, with receptor expression controlled by a tetracycline-sensitive repressor. After induction of MOPr expression with tetracycline, specific binding of [3H]DAMGO in cells expressing MOPr-WT was similar to cells expressing MOPr-L85I, MOPr-R181C, MOPr-R260H and MOPr-R265H (P > 0.05), however in cells transfected with MOPr-S268P, receptor expression was significantly reduced. Bmax for CHO-MOPr-WT cells was 280 ± 20 fmol/mg total protein and Bmax for

CHO-MOPr-S268P was 175 ± 18 fmol/mg (P < 0.05, Table 1). KD was similar between cells expressing MOPr-WT and all other variants with the exception of MOPr-R181C, which had a KD of 2.43 ± 0.53 nM compared with MOPr-WT KD of 0.75 ± 0.05 (P < 0.05,

Table 1). As CHO-MOPr-R181C and CHO-MOPr-S268P displayed aberrant DAMGO binding profiles, we have not at this stage conducted further assays using these cell lines.

Adenylyl Cyclase Inhibtion

We have previously reported opioid-mediated inhibition of FSK-stimulated membrane hyperpolarisation in CHO-MOPr cells, and this effect was naloxone and pertussis-toxin sensitive (Knapman et al., 2014). In CHO-MOPr cells loaded with membrane potential

240 dye, addition of the adenylyl cyclase activator forskolin (FSK, 300 nM) produced a hyperpolarisation in CHO-MOPr-WT of 41 ± 1 % decrease in fluorescence from baseline.

FSK stimulated a similar membrane hyperpolarisation in all cell lines (P > 0.05, Table 2).

The maximal inhibition of the FSK response by DAMGO in CHO-MOPr-WT cells was 64

± 2 %, with pEC50 of 7.9 ± 0.1 (Table 3). DAMGO inhibited FSK similarly in cells expressing MOPr-L85I, MOPr-R260H and MOPr-R265H (Table 3, Figure 1).

We have reported previously that the effects of DAMGO in this assay were sensitive to naloxone and strongly inhibited by overnight pretreatment with pertussis toxin (Knapman et al., 2014). Application of opioids alone did not affect the membrane potential of CHO-

MOPr cells, with the exception of 30 µM pentazocine and 10 µM methadone, which caused small, transient increases in fluorescence (< 10%). This effect was not naloxone sensitive.

Opioid-mediated ERK1/2 Phosphorylation

We have previously shown that in CHO cells expressing MOPr, application of opioids stimulates ERK1/2 phosphorylation, and this response is blocked by naloxone and pertussis-toxin (Knapman et al, In Press). Opioid ERK1/2 phosphorylation was normalised against application of 100 nM PMA for 10 min. The average PMA response in cells expressing MOPr-WT was 0.60 ± 0.07 (corrected absorbance reading), and did not differ significantly between cell lines (Table 2). Maximal DAMGO-stimulated ERK1/2 phosphorylation was 64 ± 3% of the PMA response in CHO-MOPr-WT cells, with a pEC50 of 7.9 ± 0.1 (Figure 1, Table 4).

241

Table 1: Surface receptor expression and DAMGO Kd for MOPr variants in CHO cells

CHO-MOPr B max K (nM) Variant (fmol/mg) d Wild Type 280 ± 20 0.75 ± 0.05 L85I 268 ± 69 0.56 ± 0.09 R181C 245 ± 31 2.43 ± 0.53* R260H 249 ± 45 1.09 ± 0.46 R265H 322 ± 49 0.76 ± 0.37 S268P 175 ± 18* 0.54 ± 0.18

Radioligand binding was carried out as described in the Methods. No significant difference in Bmax or Kd was observed between cells expressing MOPr-WT, MOPr-L85I, R260H or R265H (P > 0.05). Cells expressing MOPr-R181C had a significantly different Kd compared to MOPr-WT, and cells expressing S268P expressed significantly less receptor (P < 0.05). Further assays were not conducted on these cell lines. The assay was repeated 3 times in triplicate.

Table 2: Forskolin and PMA responses in MOPr variants

300nM FSK response 100nM PMA response CHO-MOPr (Δ fluorescence from (Corrected Abs) Variant baseline (%)) Wild Type 42 ± 1 0.60 ± 0.07 L85I 39 ± 1 0.54 ± 0.06 R260H 37 ± 3 0.59 ± 0.06 R265H 39 ± 2 0.56 ± 0.05

MOPr variants did not differ in the response to 300 nM forskolin which was used to stimulate adenylyl cyclase activity (P > 0.05). ERK1/2 phosphorylation stimulated by 100 nM PMA was similar in MOPr-WT and MOPr variants (P > 0.05). Data is from 6

242 independent experiments performed in duplicate.

Because of the possibility of ligand-specific and/or pathway specific effects of MOPr polymorphisms, we compared the potency and efficacy of a range of clinically important and/or structurally distinct opioid ligands for both AC inhibition and ERK1/2 phosphorylation in CHO cells expressing MOPr-WT and the MOPr-L85I, MOPr-R260H and MOPr-R265H variants.

MOPr-L85I

In assays of AC inhibition, DAMGO inhibition of the FSK-stimulated response in CHO-

MOPr-L85I cells was significantly higher than in CHO-MOPr-WT, with Emax of 77 ± 5%

(P < 0.05, Figure 1, Table 3). DAMGO potency was not affected by the MOPr-L85I variant, with pEC50 of 7.7 ± 0.1 (P > 0.05, Table 3). FSK inhibition by the commonly prescribed opioid analgesics morphine, buprenorphine and pentazocine was not significantly different from MOPr-WT at MOPr-L85I (Figure 1, Table 3). At concentrations higher than 10 μM, pentazocine has been reported to block K+ channels

(Zhang & Cuevas, 2005), so concentrations above 10 μM were not tested despite pentazocine response not reaching a clear plateau at this concentration. For this reason, it was not possible to calculate pEC50 for pentazocine AC inhibition for any variant.

The endogenous opioid β-endorphin inhibited FSK-stimulated hyperpolarisation similarly in cells expressing MOPr-WT or MOPr-L85I. In contrast, signalling of the endogenous opioids met-enkephalin, endomorphin-1 and endomorphin-2 was significantly enhanced in cells expressing MOP-L85I compared with cells expressing MOPr-WT in assays of AC inhibition (Figure 2, Table 3). For met-enkephalin, Emax was 63 ± 3% in CHO-MOPr-WT cells, and 99 ± 6% in CHO-MOPr-L85I cells (P < 0.05). Met-enkephalin was also more potent at MOPr-L85I, with pEC50 of 8.0 ± 0.1, compared with 7.4 ± 0.1 at MOPr-WT (P <

243 0.05). Endomorphin-1 and endomorphin-2 signalling via AC inhibition was similarly enhanced in CHO-MOPr-L85I cells (Figure 2, Table 3).

Figure 1: DAMGO, morphine, buprenorphine and pentazocine inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr variants. Adenylyl cyclase inhibition and levels of ERK1/2 phosphorylation were determined as described in the Methods. A) DAMGO, morphine, buprenorphine and pentazocine inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT and MOPr- variants to varying degrees and with different potencies (Table 3). B) DAMGO, morphine, buprenorphine and pentazocine stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-variants to varying degrees and with different potencies (Table 4). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

244 A B

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 WT 80 WT 80 L85I L85I 60 R260H 60 R260H R265H R265H 40 40

20 20

0 0 -9 -8 -7 -6 -5 -10 -9 -8 -7 -6 [DAMGO] log M [DAMGO] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 WT 80 WT 80 L85I L85I R260H 60 R260 60 R265H R265 40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Morphine] log M [Morphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 WT WT 80 80 L85I L85I 60 R260H 60 R260H R265H R265H 40 40

20 20

0 0 -10 -9 -8 -7 -6 -11 -10 -9 -8 -7 [Buprenorphine] log M [Buprenorphine] log M

Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 WT WT 80 80 L85I L85I R260H 60 R260H 60 R265H R265H 40 40

20 20

0 0 -8 -7 -6 -5 -8 -7 -6 -5 [Pentazocine] log M [Pentazocine] Log M

245

Figure 2: Endogenous opioids inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr variants. Adenylyl cyclase inhibition and levels of ERK1/2 phosphorylation were determined as described in the Methods. A) β-endorphin, met-enkephalin, endomorphin-1 and endomorphin-2 inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT and MOPr-variants to varying degrees and with different potencies (Table 3). B) β-endorphin, met-enkephalin, endomorphin-1 and endomorphin-2 stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-variants to varying degrees and with different potencies (Table 4). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

246 A B Inhibition of FSK response pERK1/2 (% Control) 100 (%) 100 WT WT 80 80 L85I L85I R260H 60 R260H 60 R265H R265H 40 40

20 20

0 0 -8 -7 -6 -5 -9 -8 -7 -6 -5 [b-Endorphin] log M [b-Endorphin] log M

Inhibition of FSK response pERK1/2 (% Control) 120 (%) 120 WT 100 WT 100 L85I L85I 80 R260H 80 R260H R265H R265H 60 60

40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Met-Enkephalin] log M [Met-Enkephalin] log M

Inhibition of FSK response pERK1/2 (% Control) (%) 100 100 WT WT 80 L85I 80 L85I R260H R260H 60 60 R265H R265H 40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-1] log M [Endomorphin-1] log M

Inhibition of FSK response pERK1/2 (% Control) 120 (%) 120

100 WT 100 WT L85I L85I 80 80 R260H R260H 60 R265H 60 R265H

40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Endomorphin-2] log M [Endormorphin-2] log M

247 We then challenged cells with the clinically important opioids methadone, fentanyl and oxycodone. The inhibition of FSK-stimulated membrane hyperpolarisation was significantly increased in cells expressing MOPr-L85I compared with MOPr-WT.

Maximal inhibition of the FSK response by methadone was 65 ± 3% in CHO-MOPr-WT cells, with pEC50 of 7.0 ± 0.1. In CHO-MOPr-L85I cells, maximal methadone inhibition of the FSK-response was 112 ± 7% (P < 0.05), resulting in a slight membrane depolarisation from baseline (Figure 3). Methadone potency was unaffected by MOPr-L85I, with pEC50 of 7.1 ± 0.1 (P > 0.05, Figure 3, Table 3). The maximally effective concentration of fentanyl also caused a slight membrane depolarisation from baseline in CHO-MOPr-L85I cells (Figure 4). MOPr-L85I Emax for fentanyl inhibition of FSK-stimulated membrane hyperpolarisation was 110 ± 9%, compared with MOPr-WT Emax of 70 ± 6% (P < 0.05,

Figure 3, Table 3). Fentanyl potency was similar between MOPr-WT and MOPr-L85I.

Oxycodone signalling was not affected by L85I (Table 3).

In assays of ERK1/2 activation, maximal DAMGO stimulation of ERK1/2 phosphorylation was not significantly different between cells expressing MOPr-WT and MOPr-L85I, with

Emax of 74 ± 6% of the maximum PMA response in CHO-MOPr-L85I cells, and pEC50 of

8.6 ± 1 (Figure 1, Table 4). The ERK1/2 phosphorylation elicited by morphine and buprenorphine was not affected by the MOPr-L85I variant, however maximum ERK1/2 phosphorylation stimulated by pentazocine was significantly decreased in cells expressing

MOPr-L85I compared with MOPr-WT. Emax for pentazocine in CHO-MOPr-WT cells was

44 ± 11%, and 16 ± 3% in CHO-MOPr-L85I cells (P < 0.05, Figure 1, Table 4).

Pentazocine pEC50 was unaffected by MOPr-L85I (Table 4).

248 Interestingly, the effect of the L85I variant on ERK1/2 signalling via endogenous opioid ligands was a mirror image of the effects observed for AC inhibition. β-endorphin- stimulated ERK1/2 phosphorylation was significantly increased at MOPr-L85I, with Emax of 99 ± 4% compared with Emax of 72 ± 5% in CHO-MOPr-WT cells (P < 0.05, Figure 3,

Table 4), although β-endorphin potency was decreased at MOPr-L85I, with pEC50 of 6.9 ±

0.1, and 7.7 ± 0.1 at MOPr-WT (P < 0.05, Table 4). ERK1/2 phosphorylation stimulated by endomorphin-1, endomorphin-2, and met-enkephalin was not affected by MOPr-L85I

(Figure 2, Table 4). Methadone, fentanyl and oxycodone ERK1/2 signalling was similar in

CHO-MOPr-WT and CHO-MOPr-L85I cells (Figure 3, Table 4).

MOPr-R260H

In cells expressing MOPr-R260H, DAMGO inhibition of FSK-stimulated membrane hyperpolarisation was not different to that in cells expressing MOPr-WT. DAMGO Emax was 64 ± 3%, and pEC50 was 7.6 ± 0.3 (Figure 1, Table 3). Morphine and buprenorphine

AC inhibition was not affected by MOPr-R260H (Table 3). In CHO-MOPr-R260H cells, pentazocine signalling was compromised (Figure 1). It was not possible to fit a standard concentration-response curve to the data, as low pentazocine concentrations inhibited AC to a similar degree as maximum pentazocine concentrations.

β-endorphin inhibition of the FSK response was significantly decreased in CHO-MOPr-

R260H cells compared with CHO-MOPr-WT cells. Maximal β-endorphin-stimulated AC inhibition was 70 ± 4% in CHO-MOPr-WT cells, and 34 ± 3% in CHO-MOPr-R260H cells (P < 0.05, Figure 3, Table 3). β-endorphin potency was not different between cells expressing MOPr-WT or MOPr-R260H (Table 3). FSK inhibition by endomorphin-1, endomorphin-2 and met-enkephalin was similar between CHO-MOPr-WT cells and CHO-

MOPr-R260H cells (Figure 3, Table 3).

249

Figure 3: Methadone, fentanyl and oxycodone inhibit adenylyl cyclase and activate ERK1/2 in CHO cells expressing MOPr-WT or MOPr variants. Adenylyl cyclase inhibition and levels of ERK1/2 phosphorylation were determined as described in the Methods. A) Methadone, fentanyl and oxycodone inhibited forskolin-stimulated adenylyl cyclase hyperpolarisation of MOPr-WT and MOPr-variants to varying degrees and with different potencies (Table 3). B) Methadone, fentanyl and oxycodone stimulated ERK1/2 phosphorylation in cells expressing MOPr-WT or MOPr-variants to varying degrees and with different potencies (Table 4). Maximum ERK1/2 phosphorylation via 100 nM PMA was used as a control for pERK1/2 experiments. Data represent the mean ± s.e.m. of pooled data from 5-6 independent determinations performed in duplicate.

250 A B

Inhibition of FSK response pERK1/2 (% Control) 120 (%) 120

100 WT 100 WT L85I L85I 80 80 R260 R260H 60 R265 60 R265H

40 40

20 20

0 0 -9 -8 -7 -6 -5 -9 -8 -7 -6 -5 [Methadone] log M [Methadone] log M

Inhibition of FSK response pERK1/2 (% Control) 120 (%) 120

100 WT 100 WT L85I L85I 80 80 R260H R260H 60 R265H 60 R265H

40 40

20 20

0 0 -10 -9 -8 -7 -6 -10 -9 -8 -7 -6 [Fentanyl] log M [Fentanyl] log M

Inhibition of FSK response pERK1/2 (% Control) (%) 100 100 WT WT L85I 80 L85I 80 R260H R260H 60 60 R265H R265H 40 40

20 20

0 0 -8 -7 -6 -5 -4 -8 -7 -6 -5 [Oxycodone] log M [Oxycodone] Log M

251

FSK/Opioid addition RFU 1000

800

600

400 300 nM FSK + HBSS 300 nM FSK + 3 mMMethadone 200 300 nM FSK + 1 mMFentanyl

0 0120 240 360 480 600 720 Time (secs)

Figure 4: High concentrations of methadone and fentanyl caused membrane depolarisation in CHO-MOPr-L85I cells. The membrane potential assay was performed as described in the Methods. Raw trace showing application of 3 µM methadone or 1 µM fentanyl with 300 nM forskolin caused a small (< 10%) increase in fluorescent signal from baseline in CHO-MOPr-L85I cells. Data shown is a representative trace from a single experiment.

252 In cells expressing MOPr-R260H, maximum methadone inhibition of the FSK response was similar to cells expressing MOPr-WT, however potency was affected, with pEC50 of

7.0 ± 0.1 in CHO-MOPr-WT cells, and 5.9 ± 0.2 in CHO-MOPr-R260H cells (P < 0.05,

Figure 3, Table 3). Maximum oxycodone inhibition of FSK was significantly decreased at

MOPr-R260H, with Emax of 37 ± 9% compared with Emax of 65 ± 4% at MOPr-WT (P <

0.05). Oxycodone potency was not affected by MOPr-R260H. Fentanyl signalling was similar between variants. (Figure 3, Table 3).

For assays of ERK1/2 phosphorylation, DAMGO stimulated ERK1/2 similarly in cells expressing MOPr-R260H and MOPr-WT (Figure 1, Table 4). Morphine-stimulated

ERK1/2 phosphorylation was significantly compromised at the R260H variant, with Emax of 43 ± 7% and pEC50 of 6.1 ± 0.3 (P < 0.05, Figure 1, Table 4). Buprenorphine and pentazocine failed to elicit significant ERK1/2 phosphorylation in CHO-MOPr-R260H cells (Figure 1, Table 4). ERK1/2 signalling by the endogenous opioids was also compromised. Maximum β-endorphin-stimulation of ERK1/2 was unchanged, but pEC50 was significantly affected in cells expressing MOPr-R260H (6.6 ± 0.2, P < 0.05, Table 4).

For endomorphin-1 and met-enkephalin, Emax and pEC50 were both affected in CHO-

MOPr-R260H cells, whereas for endomorphin-2, only Emax was affected (Figure 2, Table

4). The R260H variant also had an impact on the signalling of methadone, fentanyl and oxycodone. Maximum methadone ERK1/2 phosphorylation was similar between variants, but potency was significantly affected, with pEC50 of 7.3 ± 0.3 in CHO-MOPr-WT, and

5.5 ± 0.4 in CHO-MOPr-R260H (P < 0.05, Figure 3, Table 4). Fentanyl Emax and pEC50 were both affected by R260H, and maximum oxycodone-stimulated ERK1/2 was significantly decreased, but pEC50 was unchanged in CHO-MOPr-R260H (Figure 3, Table

4).

253 Table 3: Summary of opioid efficacy and potency in assays of AC inhibition in CHO cells expressing MOPr-WT and MOPr-variants.

AC E (%) pEC inhibition max 50 Opioid WT L85I R260H R265H WT L85I R260H R265H Fentanyl 70 ± 6 110 ± 9* 65 ± 2 67 ± 3 8.0 ± 0.2 8.1 ± 0.1 6.9 ± 0.6 7.4 ± 0.2 β-Endorphin 70 ± 4 66 ± 4 34 ± 3* 65 ± 9 6.7 ± 0.1 6.9 ± 0.2 6.3 ± 1.1 6.7 ± 0.1* Morphine 66 ± 4 81 ± 8 51 ± 9 59 ± 5 7.0 ± 0.1 6.8 ± 0.1 6.7 ± 1.0 6.8 ± 0.3 Methadone 65 ± 3 112 ± 7* 58 ± 3 76 ± 7 7.0 ± 0.1 7.1 ± 0.1 5.9 ± 0.2* 6.9 ± 0.3 Oxycodone 65 ± 4 88 ± 8 37 ± 9* 52 ± 6 5.8 ± 0.3 6.0 ± 0.2 6.0 ± 0.4 5.8 ± 0.9 Endomorphin-2 65 ± 7 97 ± 4* 66 ± 5 66 ± 4 8.2 ± 0.1 8.3 ± 0.1 7.8 ± 0.3 8.0 ± 0.1 Endomorphin-1 64 ± 5 98 ± 6* 64 ± 4 62 ± 6 8.2 ± 0.1 8.6 ± 0.1* 7.9 ± 0.3 8.1 ± 0.1 DAMGO 63 ± 3 76 ± 5* 64 ± 3 66 ± 4 7.9 ± 0.1 7.7 ± 0.1 7.6 ± 0.2 7.6 ± 0.2 Met-Enkephalin 63 ± 3 99 ± 6* 60 ± 4 56 ± 5 7.4 ± 0.1 8.0 ± 0.1* 7.5 ± 0.1 7.8 ± 0.1* Pentazocine 35 ± 5 38 ± 11 20 ± 3* 18 ± 4* N/A N/A N/A N/A 34 ± 4 26 ± 3 24 ± 6 22 ± 7 8.4 ± 0.3 7.9 ± 0.3 8.4 ± 0.5 8.4 ± 0.2 Buprenorphine

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT.

Opioids with Emax significantly lower than fentanyl are highlighted in red (1-way ANOVA, followed by Dunnett’s post- hoc test corrected for multiple comparisons, P < 0.05). MOPr-variant Emax and pEC50 values significantly different from MOPr-WT are marked with * (unpaired Student’s T-test, P <0.05).

254

Table 4: Summary of opioid efficacy and potency in assays of ERK1/2 phosphorylation in CHO cells expressing MOPr-WT and MOPr-variants.

pERK1/2 Emax (%) pEC50

Opioid WT L85I R260H R265H WT L85I R260H R265H Endomorphin-2 104 ± 11 97 ± 9 55 ± 10* 88 ± 8 8.2 ± 0.2 8.3 ± 0.2 7.8 ± 0.3 8.1 ± 0.2 Oxycodone 101 ± 10 77 ± 6 43 ± 4* 65 ± 5* 6.4 ± 0.1 6.2 ± 0.1 5.5 ± 0.4 6.2 ± 0.1 Met- 100 ± 9 96 ± 8 74 ± 6* 88 ± 11 8.2 ± 0.1 8.2 ± 0.1 7.2 ± 0.3* 8.2 ± 0.1 Enkephalin Fentanyl 99 ± 11 87 ± 7 51 ± 6* 77 ± 12 8.2 ± 0.1 8.0 ± 0.1 7.2 ± 0.3* 7.9 ± 0.3 Endomorphin-1 93 ± 5 90 ± 9 51 ± 11* 94 ± 11 8.5 ± 0.3 8.2 ± 0.2 7.5 ± 0.2* 8.1 ± 0.2

88 ± 9 77 ± 9 63 ± 10 83 ± 9 7.3 ± 0.1 7.2 ± 0.2 5.5 ± 0.4* 6.8 ± 0.1* Methadone β-Endorphin 72 ± 5 99 ± 4* 63 ± 5* 80 ± 11 7.7 ± 0.1 6.9 ± 0.1* 6.6 ± 0.2* 6.8 ± 0.2* Morphine 67 ± 8 69 ± 11 36 ± 6* 65 ± 11 7.2 ± 0.3 6.7 ± 0.1 6.1 ± 0.3* 6.6 ± 0.1

DAMGO 64 ± 6 74 ± 6 73 ± 4 69 ± 10 8.6 ± 0.1 8.3 ± 0.2 7.6 ± 0.3 7.9 ± 0.1* 45 ± 11 16 ± 3* No 20 ± 3 5.7 ± 0.4 6.5 ± 0.3 No 6.0 ± 0.5 Pentazocine Response Response 39 ± 7 32 ± 6 No 29 ± 10 8.6 ± 0.4 8.4 ± 0.3 No 8.4 ± 0.4 Buprenorphine Response Response

Assays were performed as described in the methods. Opioids are listed in rank order of maximal effect at MOPr-WT.

Opioids with Emax significantly lower than endomorphin-2 are highlighted in red (1-way ANOVA, followed by Dunnett’s post-hoc test corrected for multiple comparisons, P < 0.05). MOPr-variant Emax and pEC50 values significantly different from MOPr-WT are marked with * (unpaired Student’s T-test, P <0.05).

255 MOPr-R265H

For cells expressing MOPr-R265H, inhibition of FSK-stimulated membrane hyperpolarisation was similar to cells expressing MOPr-WT for DAMGO, morphine, buprenorphine and pentazocine (Figure 1, Table 3). The signalling of β-endorphin, endormophin-1 and endomorphin-2 was also unaffected by the R265H variant (Figure 2,

Table 3). For met-enkephalin, maximal FSK inhibition was similar between CHO-MOPr-

WT and CHO-MOPr-R265H, however pEC50 was affected, with values of 7.4 ± 0.1 and

7.8 ± 0.1, respectively (P < 0.05, Figure 2, Table 3). Fentanyl, methadone and oxycodone inhibition of the FSK-response was not significantly different at the R265H variant (Figure

3, Table 3).

DAMGO elicited a similar degree of ERK1/2 phosphorylation in CHO-MOPr-WT and

CHO-MOPr-R265H cells, however DAMGO was less potent at MOPr-R265H compared with MOPr-WT, with pEC50 of 7.9 ± 0.1 and 8.6 ± 0.1 respectively (P < 0.05, Figure 1,

Table 4). Likewise, the maximum morphine response was similar between variants, but morphine was significantly less potent at MOPr-R265H (Table 4). Buprenorphine and pentazocine ERK1/2 phosphorylation was not affected by the MOPr-R265H variant.

Maximal β-endorphin stimulation of ERK1/2 by cells expressing MOPr-R265H was similar to cells expressing MOPr-WT, however β-endorphin elicited ERK1/2 phosphorylation with lower potency, with pEC50 of 6.8 ± 0.2 (P < 0.05, Table 4, Figure 2).

Endomorphin-1, endomorphin-2 and met-enkephalin signalling did not differ between variants (Figure 2, Table 4). Fentanyl activation of ERK1/2 was similar at MOPr-WT and

MOPr-R265H. Methadone was less potent at R265H, and maximum oxycodone stimulation of ERK1/2 was decreased in CHO-MOPr-R265H cells (Figure 3, Table 4).

256 DISCUSSION

A number of MOPr polymorphisms have been identified in the population, yet as most are relatively rare, little is known about the clinical significance for individuals expressing the variants. In this study, we have shown that non-synonymous OPRM1 SNPs alter the signalling profile of MOPr in vitro, in a ligand-selective and/or pathway-specific manner.

The L85I variant, arising from a 253C > A nucleotide change in OPRM1, is a leucine to isoleucine amino acid substitution in the first transmembrane (TM1) domain of MOPr, and was first reported by Ravindranathan et al. (2009). The transmembrane helices of GPCRs are essential for transmitting information from the extracellular surface to the intracellular signalling domains, and may also contribute to ligand-selectivity of GPCRs (Laurila et al.,

2011; Law et al., 2000; Serohijos et al., 2011; Manglik et al. 2012). In this study, the L85I variant enhanced the signalling of the non-morphinan opioids DAMGO, met-enkephalin, endomorphin-1, endomorphin-2, fentanyl and methadone in assays of AC inhibition, but did not affect the morphinan opioids morphine, buprenorphine or oxycodone signalling.

Signalling of the large peptide β-endorphin, and the benzomorphan pentazocine was also unaffected. It seems unlikely that an increase in non-morphinan opioid affinity would directly cause a pathway-specific enhancement of MOPr signalling, with no change in potency. Furthermore, we found no difference in KD for DAMGO between MOPr-WT and

MOPr-L85I in agreement with a previous study (Ravindranathan et al., 2009). MOPr expression levels were similar between CHO cells expressing MOPr-WT and MOPr-L85I, thus the apparent increase in the efficacy of some opioids at MOPr-I85 is not likely to be due to an insufficient amount of receptor to fully inhibit AC in MOPr-WT cells. Another possibility for the ligand- and pathway-selectivity of L85I effects is that the three- dimensional structure of non-morphinan opioids may enable the L85I receptor-ligand complex to adopt a conformation that more effectively couples to effector molecules

257 involved in AC inhibition. Morphinan opioids and the large β-endorphin molecule may restrict the ability of MOPr-L85I to adopt a similar conformation. The L85I residue change is a relatively small change in terms of hydrophobicity and size, and perhaps would not be expected to result in any major alteration of protein structure or function. However, these amino acids are not always interchangeable (Brosnan & Brosnan, 2006). The two amino acids both play key, but different, roles in protein structure, with leucine found more frequently in α-helices, and isoleucine found more in β-sheets (Brosnan & Brosnan, 2006).

This has important implications for stability of the folded protein and the ultimate conformation of the receptor.

As we were unable to measure pentazocine inhibition of AC at maximal concentrations due to non-specific effects, it is unclear whether pentazocine signalling is affected by the

L85I variant. It would be of interest to examine other benzomorphans at this variant to determine any enhancement of activity. In assays of ERK1/2 phosphorylation, the enhancement of non-morphinan opioid signalling was absent, indicating that the receptor conformation elicited by non-morphinan opioids did not confer any increase in the ability of MOPr-L85I to stimulate ERK1/2 phosphorylation.

In contrast to our results, previous studies reported DAMGO and morphine signalling was not affected by the L85I variant in assays of intracellular Ca2+ release (Ravindranathan et al., 2009), and was moderately decreased in assays of cAMP accumulation at L83I (Cooke et al., 2014). Although our results differ from these observations, the non-morphinan ligands most markedly affected by the L85I variant in our study were not examined.

Interestingly, previous studies of the L85I variant or the rat ortholog L83I have demonstrated an enhanced capacity for morphine to promote internalisation of MOPr-

L85I/L83I (Cooke et al., 2014; Ravindranathan et al., 2009), possibly due to an increased

258 capacity for morphine to recruit GRK2 (Cooke et al.,2014). The L85I variant was also reported to show reduced tolerance and cAMP superactivation in response to chronic morphine treatement (Ravindranathan et al., 2009). The enhanced L85I/L83I trafficking, reduced adaptations to chronic morphine exposure and apparent decrease in signalling efficacy observed in these studies highlight the probability that distinct receptor conformations underlie each of these processes, and that ligand-specific receptor conformations may selectively enhance some signalling or regulatory processes over others. The apparent contradiction between our results and those of previous studies may also reflect differences in the available pool of effector proteins and regulatory molecules expressed in CHO cells compared with HEK293 cells (Atwood et al., 2011). It would be interesting to see if morphine is capable of promoting L85I internalisation in cell lines such as CHO, and if non-morphinan ligand signalling is enhanced in other cell lines such as

HEK-293.

The R260H and R265H variants are both arginine to histidine amino acid substitutions in

ICL3, arising from G > A SNPs at nucleotides 779 and 794, respectively. Although the

R260H and R265H variants are rare within the population, the importance of ICL3 in

MOPr signalling and regulation has prompted investigation into their functional consequences. The ICL domains of MOPr form major elements of the cytoplasmic interface between the receptor and intracellular effector proteins (Lefkowitz, 1998). The highly conserved ICL3 domain has been shown to be involved in basal and agonist- stimulated G-protein coupling, β-arrestin recruitment and contains multiple phosphorylation consensus sequences (Merkouris et al., 1996; Georgoussi et al., 1997;

Wang, 1999). In our study, the R260H variant in ICL3 of MOPr had a negative impact on the signalling of β-endorphin, pentazocine and oxycodone in assays of AC inhibition. In assays of ERK1/2 phosphorylation, signalling of all opioids was compromised with the

259 exception of DAMGO and methadone. Buprenorphine and pentazocine failed to elicit

ERK1/2 phosphorylation at MOPr-R260H. The effects of the R265H variant on MOPr signalling were less pronounced, with a small increase in the potency of β-endorphin and met-enkephalin for AC inhibition, and in DAMGO, β-endorphin and methadone potency for ERK1/2 phosphorylation (Table 2). Pentazocine Emax for AC inhibition and oxycodone

Emax for ERK1/2 phosphorylation were decreased at R265H.

Mutations within the ICL3 of MOPr have been shown to differentially affect agonist potency and efficacy (Chaipatikul et al., 2003), however previous studies examining

R260H or R265H signalling have provided inconsistent results (Knapman & Connor,

2014). In assays of cAMP accumulation, morphine potency and efficacy for inhibition of

FSK-stimulated radiolabelled cAMP accumulation was not affected by the H260 or H265 variants (Wang et al., 2001). Befort et al. (2001) also found no differences between H265 and WT in a cAMP response element (CRE) reporter gene assay. By contrast, Fortin et al.

(2010) found a decrease in potency of DAMGO, endomorphin-1 and leu-enkephalin signalling through both H260 and H265, using a different CRE reporter assay. It is difficult to directly compare the results of these studies and ours due to different cell backgrounds, levels of receptor expression and assay conditions used. It is possible that the high level of receptor expression in the cells used by Wang et al. (2001) and Befort et al. (2001) may have masked the changes in opioid inhibition of AC observed at the R260H and R265H variants in our assay. Furthermore, our results highlight the ligand-selective effects of

R260H and R265H on AC inhibition, while previous studies have only examined a limited number of opioids and may have failed to capture ligand-specific differences in variant signalling. The effects of R260H and R265H on MOPr function may also be pathway specific. In our study, the most marked changes in R260H signalling were observed in assays of ERK1/2 phosphorylation, a pathway that has not previously been examined with

260 regards to MOPr ICL3 variant signalling.

Measurement of the extent of AC inhibition and ERK1/2 phosphorylation was taken 5 min after the application of opioid, during which time significant receptor desensitisation and internalisation can occur (Connor et al., 2004). It is possible that the differences observed in R260H and R265H signalling may reflect differences in receptor regulation rather than a direct effect on receptor signalling, however R260H, R265H and WT-MOPr have previously been shown to be downregulated to a similar degree by high concentrations of

DAMGO (Befort et al., 2001). There is also some evidence that R260H and R265H may affect constitutive activity of MOPr, and although this evidence is far from conclusive, must be considered when interpreting results (Befort et al., 2001; Wang et al., 2001).

It is interesting to note that R260H and R265H had quite different effects on MOPr signalling, despite being identical amino acid substitutions in a similar region of MOPr.

The ICL3 domain has been shown to be one of the most dynamic regions in MOPr and other GPCRs (Cherezov et al., 2007; Rosenbaum et al., 2007; Serohijos et al., 2011), and conformational changes in this domain are related to the rearrangement of TM5 and TM6 upon activation (Rasmussen et al., 2011; Serohijos et al., 2011). It is conceivable that a minor change in the position of an R > H amino acid substitution on the structurally flexibile ICL3 domain could cause different conformations of MOPr, resulting in different patterns of MOPr signalling for R260H and R265H. Furthermore, TM5 and TM6 form important elements of the ligand-binding pocket, and the presence of ligand has been shown to dramatically alter the flexibility of ICL3 in MOPr (Serohijos et al., 2011;

Manglik et al., 2012). Aberrant arrangements of TM5 and TM6 resulting from ICL3 polymorphisms could potentially differentially affect the ability of ligand to bind to and activate MOPr-R260H or MOPr-R265H, which is consistent with our observations of ligand-selective changes in ICL3 variant signalling.

261 In this study, amino acid changes in MOPr arising from naturally occurring SNPs altered the signalling profile of MOPr in a ligand-selective and pathway-specific manner.

Although the variants described here are rare within the population, a high number of naturally occurring SNPs have been described (Kazius et al., 2008; Lotsch & Geisslinger,

2005; Ravindranathan et al., 2009), and thus a significant proportion of the population may carry one or more MOPr SNPs. All MOPr variants we examined resulted in alterations to

MOPr signalling, suggesting that MOPr function is finely regulated by the composition of amino acids and ultimate conformation of the receptor. It is thus likely that other SNPs found within the population have a high potential to affect receptor function. Presuming the variant receptors behave similarly in a physiological context, this may go some way towards explaining the individual variability in response to opioid analgesics observed within the population. As most individuals would be heterozygous for MOPr SNPs, the physiological significance of changes in MOPr variant signalling in vitro is unclear, and as the rarity of most MOPr SNPs within the population makes clinical studies impractical, further studies in cells expressing both WT and variant MOPr would be of interest.

Nevertheless, MOPr SNPs clearly have the potential to alter receptor function, and a clearer understanding of the effects of these variants could provide valuable insights into the function and regulation of MOPr. Ultimately, knowledge of the effect of individual

MOPr genotype on the response to a range of opioid analgesics could one day contribute to a more personalised approach to the prescription of opioid analgesic therapy, resulting in maximal therapeutic benefits with reduced risk of adverse effects.

262 Chapter 8

General Discussion

In this thesis, the functional consequences of five non-synonymous, naturally occurring

OPRM1 SNPs were examined. The results presented show that each MOPr variant examined caused significant alterations in MOPr signalling, and these changes were ligand-selective and pathway-specific. This supports the notion that naturally occurring polymorphisms of MOPr contribute to the individual variability observed clinically in response to opioid analgesics, and that differences in response may only be manifest for a subset of opioid drugs and/or opioid effects.

The common N40D variant on the extracellular N-terminal domain of MOPr caused a selective impairment of buprenorphine signalling for all effectors examined. The A6V variant, also on the N-terminal domain of MOPr, resulted in a complete abolition of buprenorphine signalling via AC inhibition and ERK1/2 stimulation, and a significant decrease in the efficacy and/or potency of a number of other opioids. It is not immediately apparent how variation in the N-terminal domain of MOPr causes the significant and ligand-selective effects on opioid signalling observed. The N-terminal domain does not appear to form an integral part of the ligand-binding pocket of MOPr (Manglik et al.,

2012), however, as this region was not included in the recently solved MOPr crystal structure, its contribution to the conformation of the ligand-binding pocket and the receptor as a whole is not yet understood. The N-terminus contains 5 putative glycosylation sites, one of which is removed by the N40D variant (Singh et al., 1997). This has been associated with reduced receptor expression and mRNA stability (Huang et al., 2012;

Mague et al., 2009; Wang et al., 2012; Zhang et al., 2005), yet in our cell lines MOPr-

263 N40D and MOPr-WT surface receptor expression levels were equivalent. Moreover, the

A6V variant does not affect a glycosylation site, suggesting that N-terminal polymorphisms can affect receptor function independently of glycosylation.

An N-terminal polymorphism equivalent to MOPr-A6V in the 5-HT2B receptor, R6G, increased constitutive and agonist-stimulated activity (Belmer et al., 2014), and conformational changes in the N-terminal domain of MOPr have been shown following receptor activation (Gupta et al., 2007; Gupta et al., 2008). Furthermore, truncation, point mutations of phosphorylation sites or β-arrestin binding in the C-terminal domain of MOPr affect the conformation of the N-terminus (Gupta et al., 2008). The reports from these studies and the results presented in this thesis indicate that conformation of the N-terminal and C-terminal domains are mutually dependent, and that changes to either of these regions have the potential to globally alter receptor conformation and function. Changes in the N-terminal region of other GPCRs such as the dopamine D2 and D3 receptors have been shown to affect receptor trafficking and internalisation (Cho et al., 2012; Dong &

Wu, 2006; Langelaan et al. 2013), and assays investigating the consequences of MOPr N- terminal variants on these aspects of MOPr function would be of interest.

Buprenorphine was more markedly affected by N-terminal polymorphisms compared with other opioids, and appeared to be affected in a pathway-specific manner at N40D and

A6V. Buprenorphine efficacy was significantly decreased in assays of AC inhibition and

ERK1/2 phosphorylation in CHO-MOPr-N40D cells, yet showed a decrease in potency for

GIRK activation in AtT20-MOPr-N40D cells, with no change in efficacy. Buprenorphine did not inhibit AC or stimulate ERK1/2 in CHO-MOPr-A6V, but activated GIRK channels to a similar degree as MOPr-WT in AtT20 cells, suggesting a functional selectivity towards GIRK activation, or some effect of the expression system being used.

264 Buprenorphine is a partial MOPr agonist and is an antagonist at KOPr (Pawlak et al.,

2014). One study has demonstrated pathway-specific effects of buprenorphine at MOPr

(Zaki et al., 2000). Buprenorphine showed partial agonist activity for the inhibition of cAMP and stimulation of [35S]GTPγS binding in HEK-293 cells, yet caused an increase in surface MOPr expression after treatment, similar to the increased surface receptor expression observed after treatment with the MOPr antagonist naloxone (Zaki et al., 2000).

Although significant post-activation trafficking would probably not occur within the time- frame of our assays, the results from the present study and previous reports suggest that buprenorphine may show functional selectivity towards certain signalling pathways.

Measuring ERK1/2 phosphorylation in AtT20-MOPr cells could help to determine the presence of functional bias. Examination of the signalling of closely related buprenorphine derivatives (e.g. Cami-Kobeci et al., 2011; Neilan et al., 2004), and biologically active buprenorphine metabolites (Brown et al., 2011) could also provide valuable insight into the involvement of different structural elements of the ligand or receptor in the functional changes observed at MOPr-N40D.

In addition to the abolition of buprenorphine signalling, the A6V variant had a significant impact on the signalling of a range of other opioid analgesics such as morphine and fentanyl, and on endogenous opioids including β-endorphin and endomorphins 1 and 2.

There does not appear to be a clear pattern in the opioids affected by A6V, with DAMGO, morphine and β-endorphin affected in AC inhibition assays, all opioids with the exception of β-endorphin and pentazocine affected in ERK1/2 phosphorylation assays, and morphine selectively affected in GIRK activation assays. One of the limitations of assays measuring downstream signalling pathways such as AC inhibition is that there may be significant signal amplification between ligand binding and the final measured response (Hill et al.,

2001). Ligands known to have partial agonist activity in other assay formats may manifest

265 as full agonists in cAMP accumulation assays (Hill et al., 2010). For example, in this study morphine showed full agonist activity for AC inhibition, despite being classed as a partial agonist at MOPr (Pawson et al., 2013). The differences observed between assays of AC inhibition and ERK1/2 activation may be due to a loss of sensitivity in AC inhibition assays due to signal amplification. Assays investigating different effectors in CHO cells and in particular ERK1/2 assays in AtT20 cells could help to further understand whether the differences observed between assays in ligand signalling are due to functional selectivity, assay parameters or the expression system used.

The TM1 L85I variant selectively enhanced the signalling of non-morphinan opioids for

AC inhibition, resulting in a maximum FSK inhibition that exceeded the maximum stimulated by full agonists at MOPr-WT. Receptor expression levels of MOPr-WT and

MOPr-L85I were equivalent, indicating either an increased efficiency of G protein activation by L85I, or activation of different G protein subtypes. Opioid receptors show ligand-specific patterns of Gα subtype activation (Alves et al., 2004; Massotte et al., 2001;

Sanchez-Blazquez et al., 2001; Saidak et al., 2006), which can lead to differential coupling to effector and regulatory molecules (Abrol et al., 2013). CHO cells express AC isozymes

VI and VII (Varga et al., 1998). Opioids inhibit AC VI but stimulate AC VII (Avidor-

Reiss et al., 1997), thus it seems unlikely that the increased inhibition of AC by MOPr-

L85I is caused by differential coupling to AC isozymes. AC VI is inhibited by protein kinase C as well as Gα proteins (PKC; Sadana & Dessauer, 2009). PKC is involved in many MOPr regulatory pathways (Law et al., 2000), and has been suggested to mediate uncoupling and desensitisation of the morphine-activated receptor (Bailey et al., 2009;

Kelly et al., 2008). Activation of MOPr by other opioid agonists such as DAMGO results in β-arrestin recruitment and subsequent receptor endocytosis, whereas the PKC-mediated

MOPr phosphorylation does not induce internalisation (Williams et al., 2013). Related

266 GPCRs such as 5-HT2A and dopamine receptors require PKC for endocytosis stimulated by specific ligands, indicating that PKC activation is capable of promoting internalisation

(Krillov et al., 2011; Raote et al., 2013). Altered coupling of MOPr-L85I to PKC could lead to changes in AC inhibition, and conceivably contribute to the differences in morphine-stimulated MOPr-L85I internalisation observed in previous studies (Cooke et al.

2014; Ravindranathan et al., 2009). The slight depolarisation observed with high concentrations of fentanyl and methadone requires further examination, but may be a reflection of the enhanced activity of L85I.

The R260H variant on ICL3 of MOPr decreased the ability of β-endorphin, oxycodone and pentazocine to inhibit AC, and had a negative impact on ERK1/2 phosphorylation by most opioids tested with the exception of DAMGO and methadone. The effects of the R265H variant on MOPr signalling were more subtle, with minor changes in the signalling of select opioids including β-endorphin, oxycodone and pentazocine. In agreement with these results, previous studies examining R260H and R265H signalling have reported unchanged or minor differences in DAMGO and morphine inhibition of cAMP accumulation (Befort et al., 2001; Wang et al., 2001). Fortin et al. (2010) reported decreased DAMGO, endomorphin-1 and leu-enkephalin potency using a different assay of cAMP accumulation.

These studies did not examine many of the ligands most markedly affected at

R260H/R265H in the present study, e.g. buprenorphine, highlighting the necessity for examining a range of opioids and signalling pathways to capture ligand- and pathway- specific effects of GPCR SNPs (Knapman & Connor, 2014).

The assays used here to examine differences in MOPr variant functional profiles focused on acute signalling pathways, however MOPr also undergoes receptor trafficking and longer-term regulatory and adaptive processes (Williams et al. 2013), as well as

267 constitutive or ligand-independent signalling (Connor & Traynor, 2010). The ICL3 domain is of primary importance in the coupling of MOPr to effector and regulatory molecules, contains several consensus phosphorylation sites, and serves as a docking site for accessory proteins such as β-arrestin (Lefkowitz, 1998). Befort et al. (2001) and Wang et al. (2001) both reported a decrease in constitutive GTPγS binding for R260H and R265H

MOPr variants. The ICL domains have been shown to be involved in stabilisation of the inactive state of Class A GPCRs in the absence of ligand (Chee et al., 2008; Mokrosinksi et al., 2012; Piechowski et al., 2013). MOPr ICL3 SNPs may either stabilize active conformations of the unbound receptor more readily, or enable the ligand-bound receptor to couple more effectively to effector and regulatory molecules (Rana et al., 2001).

Constitutively active GPCRs have been reported to undergo receptor desensitisation and other adaptive processes in the absence of agonist (Barak et al., 2001; Wilbanks et al.,

2002). Although the marked decrease in ligand-stimulated MOPr-R260H signalling suggests a decreased capability for ligand-stimulated MOPr effector coupling, it is possible that the differences observed for R260H and R265H in assays of “acute” signalling in the present study reflect a certain level of receptor desensitisation arising from differences in constitutive activity. It would certainly be of interest to investigate basal signalling and desensitisation of R260H and R265H to provide further insight into the functional consequences of these SNPs.

It is becoming increasingly apparent that MOPr pharmacology must be considered to be system- and context-dependent. MOPr is expressed in a wide variety of human cell types, and subtle changes in MOPr signalling arising from SNPs may generate distinct signals in different tissues and cell types, reflecting differential binding and stabilisation of distinct conformations of the receptor, varying efficacy of the receptor to couple to different pools of effector and regulatory proteins, and context-specific tissue response to activated

268 signalling pathways (Kenakin & Miller et al., 2010). For example, [35S]GTPγS binding was significantly decreased at MOPr-N40D in the secondary somatosensory area of post- mortem human brain, but was unchanged in the thalamus, despite equivalent receptor expression levels for N40D and WT (Oertel et al., 2009). It is probably necessary to examine MOPr polymorphisms in a variety of heterologous expression systems using a wide range of structurally distinct opioid ligands in order to capture the range of signalling and regulatory differences that may be produced by MOPr variants (Charfi et al., 2013).

However, examining as many effectors as possible under similar conditions enables direct comparisons between MOPr variants and/or the effects of multiple ligands. In this project, the ability of MOPr variants to inhibit AC and stimulate ERK1/2 were both examined in

CHO cells. In general, changes in signalling arising from MOPr variants were more marked in ERK1/2 assays compared with AC, and a higher proportion of opioids tested showed full agonist activity for AC inhibition than ERK1/2 phosphorylation. This probably reflects the relatively low receptor occupancy required for AC inhibition compared with ERK1/2 phosphorylation, particularly at the sub-maximal concentration of

FSK used (Nickolls et al., 2011). The use of a higher concentration of FSK may increase the sensitivity of the AC assay. It would be interesting to observe the effects of a higher

FSK concentration on AC inhibition at MOPr-L85I, to help understand the mechanism behind the enhanced AC inhibition compared with MOPr-WT. An unusual finding of this study was the relatively low efficacy of DAMGO in assays of ERK1/2 phosphorylation.

For MOPr-WT, maximum DAMGO-stimulated ERK1/2 phosphorylation was approximately 65% of the most efficacious agonist, endomorphin-2. In contrast, DAMGO shows full agonist activity in most published assays. One explanation for our results is that

DAMGO may desensitize more rapidly or reach maximum effect more slowly compared with some other agonists tested. All measurements were taken at 5 min, thus further investigation into the ideal measurement time-point for every agonist tested may be of

269 benefit. Furthermore, the difference between ligand and MOPr-variant effects on AC inhibition and ERK1/2 phosphorylation may also be related to the relative amounts of G protein subtypes in CHO cells. GIRK activation was measured in AtT-20 cells as CHO cells do not express native GIRK channels. Overall, MOPr variants showed less effect on

GIRK activation than AC inhibition and ERK1/2 activation, and this was particularly apparent with buprenorphine signalling at MOPr-A6V. The inability of buprenorphine to elicit AC inhibition or pERK1/2 phosphorylation in CHO cells at MOPr-A6V, with an unchanged capability for GIRK activation in AtT-20 cells could reflect some pathway bias, or may be related to the different cellular backgrounds. This would be best investigated by performing the ERK1/2 phosphorylation ELISA in AtT-20 cells to see if opioid signalling is affected by MOPr-variants in a comparable way to ERK1/2 assays carried out in CHO cells. It should be noted that MOPr activation of GIRK activation showed some desensitization over the course of the assay, whereas MOPr inhibition of AC did not. AC inhibition did begin to decrease 5 min after opioid addition, and we have observed further desensitization when assay duration was increased. As both assays were continuous and measurements were taken at the peak of MOPr activity, desensitization of MOPr is not likely to have significantly affected our results.

Another area of interest at present is the identification of ligands biased towards either G- protein coupled pathways or G-protein independent pathways such as β-arrestin recruitment (Violin & Lefkowitz, 2007). There is some evidence that β-arrestin-coupled pathways mediate the adverse effects arising from MOPr stimulation such as tolerance and dependence (Raehal & Bohn, 2011; Raehal et al., 2005), and current efforts to design improved opioid drugs are often focused on developing ligands with significant bias towards G-protein coupled pathways (DeWire et al., 2013). MOPr polymorphisms could also in theory show preferential coupling towards G-protein coupled pathways or G-

270 protein independent pathways, but to date this has not been explored. The assays used in this project were predominantly focused on G-protein coupled pathways, although ERK1/2 phosphorylation can occur via a number of mechanisms including β-arrestin signalling

(Law et al., 2000). Examination of the relative ability of MOPr variants to couple to β- arrestin and G proteins could provide important insights into the structural components involved in these processes as well as the mechanisms behind the pathway-specific effects observed. Defining receptor reserve using irreversible antagonists such as β- funaltrexamine or β-chlornaltrexamine and then fitting data to operational models (e.g.

Borgland et al., 2003; Rivero et al., 2012; Kelly, 2013) would also enable the precise determination of rank orders of agonist efficacy and uncover differences in signalling across different effectors in the same cell, enabling a more complete characterisation of the consequences of changes in receptor sequence (Kenakin & Christopoulos, 2013).

One area that has remained largely unexplored is the way in which variant MOPr may interact with WT-MOPr, even though most carriers of variant MOPr alleles will be heterozygous for the wild-type receptor. The MOPr crystallises as parallel dimers tightly associated through TM5 and TM6, and to a lesser extent TM1 and TM2 (Manglik et al.,

2012). Ravindranathan et al. (2009) co-expressed the L85I variant with MOPr-WT in

HEK-293 cells and reported internalisation of both L85I- and WT-MOPr in response to morphine, suggesting the formation of functional MOPr-WT/L85I dimers with L85I showing domainance. Conversely, co-expression of R181C- and WT-MOPr resulted in independent internalisation of MOPr-WT in response to DAMGO, indicating R181C either does not dimerise with WT, or that the interaction is unstable (Ravindranathan et al.,

2009). The effects of co-expression of these variants on MOPr signalling were not investigated. Other functional studies of MOPr SNPs have not examined the possible interactions between WT- and variant-MOPr, the formation of homo- or heterodimers, or

271 the effects of co-expressed MOPr variants on cellular signalling. Although there is growing evidence supporting the ability of MOPr and other GPCRs to form homo- or heterodimers under experimental conditions (Golebiewska et al., 2011; Jordan & Devi, 1999; Juhasz et al., 2008; Milligan, 2009; Zheng et al., 2012), whether this occurs in a physiological setting is still a matter of debate (Chabre & LeMaire, 2005; Herrick-Davis et al., 2013;

James et al., 2006; Kuscak et al., 2009; Meyer et al., 2006). In one study, human MOPr was reconstituted in monomeric form following expression in insect cells, and the monomer was shown to be capable of ligand-binding, G protein activation and allosteric modulation (Kuszak et al., 2009), indicating that the formation of oligomers is not necessary for MOPr function. On the other hand, a study using a novel fluorescent correlation spectroscopy technique suggested that freely diffusing class A GPCRs exist primarily as homodimers (Herrick-Davis et al., 2013). In any case, the consequences of co-expression of WT and mutant MOPr alleles is certainly an area requiring further investigation, although limited by the technical difficulty of accurately quantifying and expressing equivalent amounts of variant receptor (Liu et al., 2000). The use of bicistronic vectors, which allow the expression of 2 genes using a single promoter, should in theory result in equal expression of WT-MOPr and variant MOPr, although this is not always the case in practice (Kim et al., 2004b; Pfutzner, 2008). A recent study suggested that insert size may cause the differences in relative transfection efficiency and expression of the two inserts in bicistronic vectors (Payne et al., 2013). As MOPr SNP variants are identical in length, this method may be appropriate for future examination of co-expressed MOPr variants.

The N40D variant is present in up to 50% of individuals in some populations (Mura et al.,

2013), while the A6V variant is present in upwards of 20% of individuals from other populations (Crowley et al., 2003; Kapur et al., 2007). Although other identified MOPr

272 SNPs are rare (< 1%; Lotsch & Geisslinger 2005), a substantial proportion of the population may carry one or more MOPr variants. Each MOPr variant examined in this study had some impact on MOPr signalling, indicating that the ultimate conformation of

MOPr is highly sensitive to changes in the amino acid sequence, and that this has the potential to significantly alter receptor signalling and/or regulation. Opioid analgesics remain the mainstay of analgesic therapy, despite the associated adverse effects and the potential for tolerance and dependence to develop. The total number of opioid prescriptions in Australia in 2010 exceeded 11 million (Mabbott et al., 2010).

Buprenorphine accounted for almost 10% of total opioid prescriptions, exceeded only by , and oxycodone (Mabbott et al., 2010). In the present study, buprenorphine signalling was significantly affected by the N40D, A6V and R260H MOPr variants. Other commonly prescribed opioids including oxycodone (24% of total prescriptions) and morphine (6% of total prescriptions, 40% including codeine; Mabbott et al., 2010) were less effective at the A6V variant. Assuming MOPr variants behave similarly in a neuronal context, a significant proportion of the population may be receiving inappropriate analgesic therapy, and could potentially benefit from prescription of an alternative opioid analgesic. Understanding the consequences of expressing a particular

MOPr variant should in the future enable the development of personalized prescription plans depending on individual phenotype. Individual opioid response involves multiple proteins involved in opioid ligand distribution and metabolism, as well as effectors downstream of MOPr, which would also be affected by genotype. Neverthelss, knowledge of an indiviudal’s MOPr genotype would be a key element in the ability to predict the effects of specific opioid drugs, including side effects and the development of tolerance, minimizing the risk of serious adverse events associated with opioid overdose, while maximizing therapeutic benefits and ensuring individuals receive adequate pain relief.

273

274 REFERENCES

Abrol R, Kim SK, Bray JK, Trzaskowski B, Goddard WA 3rd (2013). Conformational ensemble view of G protein-coupled receptors and the effect of mutations and ligand binding. Methods Enzymol 520: 31-48.

Akil H, Watson SJ, Young E, Lewis ME, Khachaturian H, Walker JM (1984). Endogenous opioids: biology and function. Annu Rev Neurosci 7: 223-255.

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Spedding M et al.

(2013). The concise guide to PHARMACOLOGY 2013/14: G protein-coupled receptors.

Br J Pharmacol 170: 1459-1581.

Alves ID, Salamon Z, Varga E, Yamamura HI, Tollin G, Hruby VJ (2003). Direct observation of G-protein binding to the human δ-opioid receptor using plasmon-waveguide resonance spectroscopy. J Biol Chem 278: 48890-48897.

Aoki T, Narita M, Ohnishi O, Mizuo K, Narita M, Yajima Y et al. (2003). Disruption of the type 1 inositol 1,4,5-triphosphate receptor gene suppresses the morphine-induced antinociception in the mouse. Neurosci Lett 350: 69-72.

Arvidsson U, Riedl M, Chakrabarti S, Lee JH, Nakano AH, Dado RJ et al. (1995).

Distribution and targeting of a mu-opioid receptor (MOR1) in brain and spinal cord. J

Neurosci 15: 3328-3341.

275 Atwood BK, Lopez J, Wager-Miller J, Mackie K, Straiker A (2011). Expression of G protein-coupled receptors and related proteins in HEK293, AtT20, BV2, and N18 cell lines as revealed by microarray analysis. BMC Genomics 12:14

Avidor-Reiss T, Nevo I, Saya D, Bayewitch M, Vogel Z (1997). Opiate-induced adenylyl cyclase superactivation is isozyme-specific. J Biol Chem 272: 5040-5047.

Bailey CP, Oldfield S, Llorente J, Caunt CJ, Teschemacher AG, Roberts L and others

(2009). Involvement of PKC alpha and G-protein-coupled receptor kinase 2 in agonist- selective desensitisation of mu-opioid receptors in mature brain neurons. Br J Pharmacol

158: 157-164.

Barak LS, Oakley RH, Laporte SA, Caron MG (2001). Constitutive arrestin-mediated desensitisation of a human mutant associated with nephrogenic diabetes insipidus. Proc Natl Acad Sci U S A 98: 93-98.

Bart G, Heillig M, LaForge KS, Pollak L, Leal SM, Ott J et al (2004). Substantial attributable risk related to a functional μ-opioid receptor gene polymorphism in association with heroin addiction in central Sweden. Mol Psychiatry 9: 547-549.

Bart G, Kreek MJ, Ott J, LaForge KS, Proudnikov D, Pollak L et al. (2005). Increased attributable risk related to a functional μ-opioid receptor gene polymorphism in association with alcohol dependence in central Sweden. Neuropsychopharmacology 30: 417-422.

Basbaum AI, Bautista DM, Scherrer G, Julius D (2009). Cellular and molecular mechanisms of pain. Cell 139: 267-284.

276 Becker A, Gresksch G, Brodemann R, Kraus J, Peters B, Schroeder H et al. (2000).

Morphine self-administration in mu-opioid receptor-deficient mice. Naunyn

Schmiedebergs Arch Pharmacol 361: 584-589.

Beckett AH, Casy AF (1954). Synthetic analgesics: stereochemical considerations. J

Pharm Pharmacol 6: 986-999.

Befort K, Filliol D, Decaillot FM, Gaveriaux-Ruff C, Hoehe MR, Kieffer BL (2001). A single nucleotide polymorphic mutation in the human mu-opioid receptor severely impairs receptor signalling. J Biol Chem 276: 3130-7.

Belmer A, Doly S, Setola V, Banas SM, Moutkine I, Boutourlinsky K et al. (2014). Role of the N-terminal region in G protein-coupled receptor functions: Negative modulation revealed by 5-HT2B receptor polymorphisms. Mol Pharmacol 85: 127-138.

Bergen AW, Kokoszka J, Peterson R, Long JC, Virkkunen M, Goldman D (1997). μ- opioid receptor gene variants: lack of association with alcohol dependence. Mol Psych 2:

490-494.

Berrettini WH, Hoehe MR, Ferraro TN, DeMaria PA, Gottheil E (1997). Human mu- opioid receptor gene polymorphisms and vulnerability to substance abuse. Addiction Biol

2: 303-308.

277 Beyer A, Koch T, Schroder H, Schulz S, Hollt V (2004). Effect of the A118G polymorphism on binding affinity, potency and agonist-mediated endocytosis, desensitisation and resensitisation of the human mu-opioid receptor. J Neurochem 89: 553-

560.

Bilecki W, Zapart G, Ligeza A, Wawrzczak-Bargiela A, Urbanski MJ, Przewlocki R

(2005). Regulation of the extracellular signal-regulated kinases following acute and chronic opioid treatment. Cell Mol Life Sci 62: 2369-2375.

Blackburn TP, Cross AJ, Hille C, Slater P (1988). Autoradiographic localisation of delta opiate receptors in rat and human brain. Neuroscience 27: 497-506.

Bond C, LaForge KS, Tian M, Melia D, Zhang S, Borg L et al. (1998). Single-nucleotide polymorphism in the human mu opioid receptor gene alters β-endorphin binding and activity: Possible implications for opiate addiction. Proc Natl Acad Sci USA 95: 9608-9613

Boom M, Niesters M, Sarton E, Aarts L, Smith TW, Dahan A (2012). Non-analgesic effects of opioids: opioid-induced respiratory depression. Curr Pharm Des 18: 5994-6004.

Borgland SL, Connor M, Osborne PB, Furness JB, Christie MJ (2003). Opioid agonists have different efficacy profiles for G protein activation, rapid desensitisation, and endocytosis of μ-Opioid receptors. J Biol Chem 278: 18776-18784.

Borland G, O Smith B, Yarwood SJ (2009). EPAC proteins transduce diverse cellular actions of cAMP. Br J Pharmacol 158: 70-86.

278 Borsodi A, Calo G, Chavkin C, Christie M, Civelli O, Cox BM et al. (2013). IUPHAR database (IUPHAR-DB), http://www.iuphar-db.org, accessed 10/02/2014.

Brock C, Olesen SS, Olesen AE, Frokjaer JB, Andresen T, Drewes AM (2012). Opioid- induced bowel dysfunction: pathophysiology and management. Drugs 72: 1847-1865.

Brosnan JT, Brosnan ME (2006). Branched-chain amino acids: enzyme and substrate regulation. J Nutr 136: 207S-211S.

Brown SM, Holtzman M, Kim T, Kharasch ED (2011). Buprenorphine metabolites, buprenorphine-3-glucuronide and -3-glucuronide, are biologically active.

Anesthesiology 115: 1251-1260.

Buntin-Mashock C, Phillip L, Moriyama K, Palmer PP (2005). Age-dependent opioid escalation in chronic pain patients. Anesth Analg 100: 1740-1745.

Bunzow JR, Saez C, Mortrud M, Bouvier C, Williams JT Low M et al. (1994). Molecular cloning and tissue distribution of a putative member of the rat opioid receptor gene family that is not a mu, delta or kappa opioid receptor type. FEBS Lett 347: 284-288.

Cami-Kobeci G, Polgar WE, Khroyan TV, Toll L, Husbands SM (2011). Structural determinants of opioid and NOP receptor activity in derivatives of buprenorphine. J Med

Chem 54: 6531-6537.

279 Campa D, Gioia A, Tomei A, Poli P, Barale R (2008). Association of ABCB1/MDR1 and

OPRM1 gene polymorphisms with morphine pain relief. Clin Pharmacol Ther 83: 559-

566.

Chabre M, LeMaire M (2005). Monomeric G-protein-coupled receptor as a functional unit.

Biochemistry 44: 9395-9403.

Chaipatikul V, Loh, HH, Law PY (2003). Ligand-selective activation of μ-opioid receptor:

Demonstrated with deletion and single amino acid mutations of third intracellular loop domain. J Pharmacol Exp Ther 305: 909-918.

Chakrabarti S, Prather PL, Yu L, Law PY, Loh HH (1995). Expression of the µ-opioid receptor in CHO cells: ability of µ-opioid ligands to promote α-azidoanilido[32P] GTP labelling of multiple G protein α subunits. J Neurochem 64: 2534-2543.

Chamorro AJ, Marcos M, Miron-Canelo JA, Pastor I, Gonzalez-Sarmiento R, Laso FJ

(2012). Association of μ-opioid receptor (OPRM1) gene polymorphism with response to naltrexone in alcohol dependence: a systematic review and meta-analysis. Addict Biol 17:

505-512.

Charbogne P, Kieffer BL, Befort K (2014). 15 years of genetic approaches in vivo for addiction research: Opioid receptor and peptide gene knockout in mouse models of drug abuse. Neuropharmacol 76: 204-217.

280 Charfi I, Nagi K, Mnie-Filali O, Thibault D, Balboni G, Schiller PW et al (2014). Ligand- and cell-dependent determinants of internalisation and cAMP modulation by delta opioid receptor (DOR) agonists. Cell Mol Life Sci 71: 1529-1546.

Chee MJS, Morl K, Lindner D, Merten N, Zamponi G, Light PE et al (2008). The third intracellular loop stabilizes the inactive state of the neuropeptide Y1 receptor. J Biol Chem

283: 33337-33346.

Chen Y, Mestek A, Liu J, Yu L (1993). Molecular cloning of a rat kappa opioid receptor reveals sequence similarities to the mu and delta opioid receptors. Biochem J 295: 625-

628.

Cho DI, Min C, Jung KS, Cheong SY, Zheng M, Cheong SJ et al. (2012). The N-terminal region of the dopamine D2 receptor, a rhodopsin-like GPCR, regulates correct integration into the plasma membrane and endocytic routes. Br J Pharmacol 166: 659-675.

Chong RY, Oswald L, Yang X, Uhart M, Lin PI, Wand GS (2006). The μ-opioid receptor polymorphism A118G predicts cortisol responses to naloxone and stress.

Neuropsychopharmacology 31: 204-211.

Chou WY, Wang CH, Liu PH, Liu CC, Tseng CC, Jawan B (2006). Human opioid receptor A118G polymorphism affects intravenous patient-controlled analgesia morphine consumption after total abdominal hysterectomy. Anaesthesiology 105: 334-337.

281 Chou WY, Yang LC, Lu HF, Ko JY, Wang CH, Lin SH et al. (2006). Association of μ- opioid receptor gene polymorphism (A118G) with variations in morphine consumption for analgesia after total knee arthroplasty. Acta Anaesthesiol Scand 50: 787-792.

Christie MJ (2008). Cellular neuroadaptations to chronic opioids: tolerance, withdrawal and addiction. Br J Pharmacol 154: 384-396.

Christie MJ, Connor M, Vaughan CW, Ingram SL, Bagley EE (2000). Cellular actions of opioids and other analgesics: implications for synergism in pain relief. Clin Exp

Pharmacol Physiol 27: 520-523.

Chu J, Zheng H, Loh HH, Law PY (2008). Morphine-induced μ-opioid receptor rapid desensitisation is independent of receptor phosphorylation and β-arrestins. Cell Signal 20:

1616-1624.

Chu J, Zheng H, Zhang Y, Loh HH, Law PY (2010). Agonist-dependent μ-opioid receptor signalling can lead to heterologous desensitisation. Cell Signal 22: 684-696.

Chung PCS, Kieffer BL (2013). Delta opioid receptors in brain function and diseases.

Pharmacol Ther 140: 112-120.

Clark JD, Tempel BL (1998). Hyperalgesia in mice lacking the Kv1.1 potassium channel gene. Neurosci Lett 251: 121-124.

Coggeshall RE, Zhou S, Carlton SM (1997). Opioid receptors on peripheral sensory axons.

Brain Res 764: 126-132.

282 Coller JK, Cahill S, Edmonds C, Farquharson AL, Longo M, Minniti R et al (2011).

OPRM1 A118G genotype fails to predict the effectiveness of naltrexone treatment for alcohol dependence. Pharmacogenet Genomics 21: 902-905.

Collier HO, Roy AC (1974). Hypothesis: inhibition of E prostaglandin-sensitive adenylyl cyclase as the mechanism of morphine analgesia. Prostaglandins 10: 361-376.

Compton P, Geschwind DH, Alarcon M (2003). Association between human mu-opioid receptor gene polymorphism, pain tolerance, and opioid addiction. Am J Med Genet B

Neuropsychiatr Genet 121B(1): 76-82.

Connor M, Borgland SL, Christie MJ (1999). Continued morphine modulation of calcium channel currents in acutely isolated locus coeruleus neurons from morphine-dependent rats. Br J Pharmacol 128: 1561-1569.

Connor M, Christie MD (1999). Opioid receptor signalling mechanisms. Clin Exp

Pharmacol Physiol 26: 493-499.

Connor M, Henderson G (1996). δ and μ-opioid receptor mobilisation of intracellular calcium in SH-SY5Y human neuroblastoma cells. Br J Pharmacol 117: 333-340.

Connor M, Kitchen I (2006). Has the sun set on kappa3-opioid receptors? Br J Pharmacol

147: 349-350.

Connor M, Osborne PB, Christie MJ (2004). Mu-opioid receptor desensitisation: is morphine different? Br J Pharmacol 143: 685-696.

283 Connor M, Schuller A, Pintar JE, Christie MJ (1999). μ-Opioid receptor modulation of calcium channel current in periaqueductal grey neurons from C57B16/J mice and mutant mice lacking MOR-1. Br J Pharmacol 126: 1553-1558.

Connor M, Traynor J (2010). Constitutively active μ-opioid receptors. Methods Enzymol

484: 445-469.

Contarino A, Picetti R, Matthes HW, Koob GF, Kieffer BL, Gold LH (2002). Lack of reward and locomotor stimulation induced by heroin in mu-opioid receptor-deficient mice.

Eur J Pharmacol 446: 103-109.

Cooke AE, Oldfield S, Krasel C, Mundell SJ, Henderson G, Kelly E (2014). Morphine- induced internalisation of the L83I mutant of the rat μ-opioid receptor. Br J Pharmacol, In

Press doi: 10.1111/bph.12709.

Corbett AD, Henderson G, McKnight AT, Paterson SJ (2006). 75 years of opioid research: the exciting but vain quest for the Holy Grail. Br J Pharmacol 147 Suppl 1: S153-162.

Corey S, Clapham DE (2001). The stoichiometry of Gβγ binding to G-protein-regulated inwardly rectifying K+ channels (GIRKs). J Biol Chem 276: 11409-11413.

Costa T, Herz A (1989). Antagonists with negative intrinsic activity at delta opioid receptors coupled to GTP-binding proteins. Proc Natl Acad Sci U S A 86: 7321-7325.

Cox B (2013). Recent developments in the study of opioid receptors. Mol Pharmacol

83:723-728.

284 Crowley JJ, Oslin DW, Patkar AA, Gottheil E, DeMaria PA Jr, O’Brien CP et al. (2003).

A genetic association study of the mu opioid receptor and severe opioid dependence.

Psychiatr Genet 13: 169-73.

Crystal HA, Hamon S, Randesi M, Cook J, Anastos K, Lazar J (2012). A C17T polymorphism in the mu opiate receptor is associated with quantitative measures of drug use in African American women. Addict Biol 17: 181-91.

Cunha FQ, Teixeira MM, Ferreira SH (1999). Pharmacological modulation of secondary mediator systems – cyclic AMP and cyclic GMP – on inflammatory hyperalgesia. Br J

Pharmacol 127: 671-678.

Cunha TM, Roman-Campos D, Lotufo CM, Duarte HL, Souza GR, Verri WA et al.

(2010). Morphine peripheral analgesia depends on activation of the

PI3Kγ/AKT/nNOS/NO/KATP signalling pathway. Proc Natl Acad Sci USA 107: 4442-

4447.

Dahan A, Aarts L & Smith TW (2010). Incidence, reversal and prevention of opioid- induced respiratory depression. Anesthesiology 112: 226-238.

DeWire SM, Yamashita DS, Rominger DH, Liu G, Cowan CL, Graczyk TM et al. (2013).

A G protein-biased ligand at the μ-opioid receptor is potently analgesic with reduced gastrointestinal and respiratory dysfunction compared with morphine. J Pharmacol Exp

Ther 344: 708-717.

285 Dhawan BN, Cesselin F, Raghubir R, Reisine T, Bradley PB, Portoghese PS et al. (1996).

International union of pharmacology XII. Classification of opioid receptors. Pharmacol

Rev 48: 567-592.

Dietis N, Rowbotham DJ, Lambert DG (2011). Opioid receptor subtypes: fact or artifact?

Br J Anaesth 107: 8-18.

Diverse-Pierluissi M, McIntire WE, Myung CS, Lindorfer MA, Garrison JC, Goy MF et al. (2000). Selective coupling of G protein βγ complexes to inhibition of Ca2+ channels. J

Biol Chem 275: 28380-28385.

Doll C, Konietzko J, Poll F, Koch T, Hollt V, Schulz S (2011). Agonist-selective patterns of mu-opioid receptor phosphorylation revealed by phosphosite-specific antibodies. Br J

Pharmacol 164: 298-307.

Dolphin AC (2003). G protein modulation of voltage-gated calcium channels. Pharmacol

Rev 55: 607-627.

Dong C, Wu G (2006). Regulation of anterograde transport of alpha2-adrenergic receptors by the N termini at multiple intracellular compartments. J Biol Chem 281: 38543-38554.

Drakenburg K, Nikoshkov A, Horvath MC, Fagergren P, Gharibyan A, Saarelainen K et al.

(2006). Mu opioid receptor A118G polymorphism in association with striatal opioid neuropeptide gene expression in heroin abusers. Proc Natl Acad Sci U S A 103: 7883-

7888.

286 Drolet G, Dumont EC, Gosselin I, Kinkead R, LaForest S, Trottier J-F (2001). Role of endogenous opioid system in the regulation of the stress response. Prog Neuro-

Psychopharmacol & Biol Psychiat 25: 729-741.

Du Y, Wan YJ (2009). The interaction of reward genes with environmental factors in contribution to alcoholism in Mexican Americans. Alcohol Clin Exp Res 33: 2103-2112.

Ducat E, Ray B, Bart G, Umemura Y, Varon J, Ho A et al (2013). Mu-opioid receptor

A118G polymorphism in healthy volunteers affects hypothalamic-pituitary-adrenal axis adrenocorticotropic hormone stress response to metyrapone. Addict Biol 18: 325-331.

Eijkelkamp N, Linley JE, Torres JM, Bee L, Dickenson AH, Gringhuis M et al. (2013). A role for Piezo2 in EPAC1-dependent mechanical allodynia. Nat Commun 4: 1682.

El Kouhen R, Burd AL, Erickson-Herbrandson, LJ, Chang CY, Law PY, Loh HH (2001).

Phosphorylation of Ser363, Thr370 and Ser375 residues within the carboxyl tail differentially regulates μ-opioid receptor internalisation. J Biol Chem 276: 12774-12780.

Evans BA, Broxton N, Merlin J, Sato M, Hutchinson DS, Christopoulos A et al. (2011).

Quantification of functional selectivity at the human α(1A)-adrenoceptor. Mol Pharmacol

79: 298-307.

Evans CJ, Keith DE Jr, Morrison H, Magendzo K, Edwards RH (1992). Cloning of a delta opioid receptor by functional expression. Science 258: 1952-1955.

287 Filliol D, Ghozland S, Chluba J, Martin M, Matthes HWD, Simonin F et al. (2000). Mice deficient for δ- and µ- opioid receptors exhibit opposing alterations of emotional responses. Nat Genet 25: 195-200.

Finnegan TF, Chen SR, Pan HL (2006). Mu opioid receptor activation inhibits GABAergic inputs to basolateral amygdala neurons through Kv1.1/1.2 channels. J Neurophysiol 95:

2032-2041.

Fortin JP, Ci L, Schroeder J, Goldstein C, Montefusco MC, Peter I et al. (2010). The μ- opioid receptor variant N190K is unresponsive to peptide agonists yet can be rescued by small-molecule drugs. Mol Pharmacol 78: 837-845.

Fuchs PN, Roza C, Sora I, Uhl G, Raja SN (1999). Characterisation of mechanical withdrawal responses and effects of mu-, delta- and kappa-opoid agonists in mornal and mu-opioid receptor knockout mice. Brain Res 13: 480-486.

Fukuda K, Hayashida M, Ikeda K, Koukita Y, Ichinohe T, Kaneko Y (2010). Diversity of opioid requirements for postoperative pain control following oral surgery – is it affected by polymorphism of the μ-opioid receptor? Anesth Prog 57: 145-149.

Galleotti N, Ghelardini C, Papucci L, Capaccioli S, Quattrone A, Bartolini A (1997). An antisense oligonucleotide on the mouse Shaker-like potassium channel Kv1.1 gene prevents antinociception induced by morphine and baclofen. J Pharmacol Exp Ther 281: 941-949.

288 Garzon J, Garcia-Espana A, Sanchez-Blazquez P (1997). Opioids binding µ- and δ receptors exhibit diverse efficacy in the activation of Gi2 and Gx/z transducer proteins in mouse matter. J Pharmacol Exp Ther 281: 549-557/

Gaveriaux-Ruff C, Karchewski LA, Hever X, Matifas A, Kieffer BL (2008). Inflammatory pain is enhanced in delta opioid receptor-knockout mice. Eur J Neurosci 27: 2558-2567.

Gaveriaux-Ruff C, Kieffer BL (2011). Delta opioid receptor analgesia: recent contributions from pharmacology and molecular approaches. Behav Pharmacol 22: 405-414.

Gaveriaux-Ruff C, Nozaki C, Nadal X, Hever XC, Weibel R, Matifas A et al. (2011).

Genetic ablation of delta opioid receptors in nociceptive sensory neurons increases chronic pain and abolishes opioid analgesia. Pain 152: 1238-1248.

Gaveriaux-Ruff C, Simonin F, Filliol D, Kieffer BL (2003). Enhanced humoral response in kappa-opioid receptor knockout mice. J Neuroimmunol 134: 72-81.

Gavioli EC, Rizzi A, Marzola G, Zucchini S, Regoli D, Calo G (2007). Altered anxiety- related behavior in nociception/orphanin FQ receptor gene knockout mice. Peptides 28:

1229-1239.

Georgoussi Z, Merkouris M, Mullaney I, Megaritis G, Carr C, Zioudrou C, Milligan G

(1997). Selective interactions of mu-opioid receptors with pertussis toxin-sensitive G proteins: involvement of the third intracellular loop and the c-terminal tail in coupling.

Biochim Biophys Acta 1359: 263-274.

289 Gerra G, Somaini L, Leonardi C, Cortese E, Maremmani I, Manfredini M et al. (2014).

Association between gene variants and response to buprenorphine maintenance treatment.

Psychiatry Res 215: 202-207.

Gether U (2000). Uncovering molecular mechanisms involved in activation of G protein- coupled receptors. Endocr Rev 21: 90-113.

Ghanouni P, Steenhuis JJ, Farrens DL, Kobilka BK (2001). Agonist-induced conformational changes in the G-protein coupling domain of the β2-adrenergic receptor.

Proc Natl Acad Sci U S A 98: 5997-6002.

Gianoulakis C (1998). Alcohol-seeking behaviour: the roles of the hypothalamic-pituitary- adrenal axis and the endogenous opioid system. Alcohol Health Res World 22: 202-210.

Glatt SJ, Bousman C, Wang RS, Murthy KK, Rana BK, Lasky-Su JA et al. (2007).

Evaluation of OPRM1 variants in heroin dependence by family-based association testing and meta-analysis. Drug Alcohol Depend 90: 159-165.

Goldsmith ZG, Dhanasekaran DN (2007). G protein regulation of MAPK networks.

Oncogene 26: 3122-3142.

Goldstein A, Tachibana S, Lowney LI, Hunkapiller M, Hood L (1979). -(1-13), an extraordinarily potent . Proc Natl Acad Sci U S A 76: 6666-6670.

290 Golebiewska U, Johnston JM, Devi L, Filizola M, Scarlata S (2011). Differential response to morphine of the oligomeric state of μ-opioid in the presence of δ-opioid receptors. Cell

Stem Cell 8: 136-147.

Gomes I, Ijzerman AP, Ye K, Maillet EL, Devi LA (2011). G protein-coupled receptor heteromerisation: a role in allosteric modulation of ligand binding. Mol Pharmacol 79:

1044-1052.

Gonzalez LG, Portillo E, Del Pozo E, Baeyens JM (2001). Changes in [(3H)]glibenclamide binding to mouse forebrain membranes during morphine tolerance. Eur J Pharmacol 418:

29-37.

Granier S, Manglik A, Kruse AC, Kobilka TS, Thian FS, Weis WI et al. (2012). Structure of the δ-opioid receptor bound to naltrindole. Nature 485: 400-404.

Grecksch G, Just S, Pierstorff C, Imhof AK, Gluck, L, Doll C et al. (2011). Analgesic tolerance to high-efficacy agonists but not to morphine is diminished in phosphorylation- deficient S375A mu-opioid receptor knock-in mice. J Neurosci 31:13890-13896.

Griffin RS (2005). An Epac-dependent pain pathway. J Neurosci 25: 8113-8114.

Gscheidel N, Sander T, Wendel B, Heere P, Schmidt LG, Rommelspacher H et al. (2000).

Five exon 1 variants of μ-opioid receptor and vulnerability to alcohol dependence. Pol J

Pharmacol 52: 27-31.

291 Gupta A, Decaillot FM, Gomes I, Tkalych O, Heimann AS, Ferro ES et al. (2007).

Conformation state-sensitive antibodies to G-protein-coupled receptors. J Biol Chem 282:

5116-5124.

Gupta A, Rozenfeld R, Gomes I, Raehal KM, Decaillot FM, Bohn LM et al. (2008). Post- activation-mediated changes in opioid receptors detected by N-terminal antibodies. J Biol

Chem 283: 10735-10744.

Hanks GW, Conno F, Cherny N, Hanna M, Kalso E, McQuay HJ et al. (2001). Morphine and alternative opioids in cancer pain: the EAPC recommendations. Br J Cancer 84: 587-

593.

Harrison C, McNulty S, Smart D, Rowbotham DJ, Grandy DK, Devi LA et al. (1999). The effects of endomorphin-1 and endomorphin-2 in CHO cells expressing recombinant mu- opioid receptors and SH-SY-5Y cells. Br J Pharmacol 128: 472-478.

Hauser KF, Stiene-Martin A, Mattson MP, Elde RP, Ryan SE, Godleske CC (1996). Mu- opioid receptor-induced Ca2+ mobilisation and astroglial development: morphine inhibits

DNA synthesis and stimulates cellular hypertrophy through a Ca2+-dependent mechanism.

Brain Res 720: 191-203.

Heinricher MM (2003). Orphanin FQ/nociception: from neural circuitry to behavior. Life

Sci 73: 813-822.

292 Hernandez-Avila CA, Covault J, Wand G, Zhang H, Gelernter J, Kranzler HR (2007).

Population-specific effects of the Asn40Asp polymorphism at the μ-opioid receptor gene

(OPRM1). Pharmacogenet Genomics 17: 1031-1038.

Hernandez-Avila CA, Wand G, Luo X, Gelernter J, Kranzier HR (2003). Association between the cortisol response to opioid blockade and the Asn40Asp polymorphism at the

μ-opioid receptor locus (OPRM1). Am J Med Genet B Neuropsychiatr Genet 118B: 60-65.

Herrick-Davis K, Grinde E, Cowan A, Mazurkiewicz JE (2013). Fluorescence correlation spectroscopy analysis of serotonin, adrenergic, muscarinic, and dopamine receptor dimerisation: the oligomer number puzzle. Mol Pharmacol 84: 630-642.

Hibino H, Inanobe A, Furutani K, Murakami S, Findlay I, Kurachi Y (2010). Inwardly rectifying potassium channels: their structure, function and physiological states. Physiol

Rev 90: 291-366.

Hill SJ, Baker JG, Rees S (2001). Reporter-gene systems for the study of G-protein- coupled receptors. Curr Opin Pharmacol 1: 526-532.

Hill SJ, Williams C, May LT (2010). Insights into GPCR pharmacology from the measurement of changes in intracellular cyclic AMP; advantages and pitfalls of differing methodologies. Br J Pharm 161: 1266-1275.

Huang P, Chen C, Mague SD, Blendy JA, Liu-Chen LY (2012). A common single nucleotide polymorphism A118G of the μ opioid receptor alters its N-glycosylation and protein stability. Biochem J 441: 379-386.

293

Hucho TB, Dina OA, Levine JD (2005). Epac mediates a cAMP to PKC signalling in inflammatory pain: an isolectin B4(+) neuron-specific mechanism. J Neurosci 25: 6119-

6126.

Hucho T, Levine JD (2007). Signalling pathways in sensitisation: toward a cell biology. Neuron 55: 365-376.

Hughes J, Smith TW, Kosterlitz HW, Fothergill LA, Morgan BA, Morris HR (1975).

Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature 258: 577-580.

Hull LC, LLorente J, Gabra BH, Smith FL, Kelly E, Bailey C et al. (2010). The effect of protein kinase C and G protein-coupled receptor kinase inhibition on tolerance induced by

μ-opioid agonists of different efficacy. J Pharmacol Exp Ther 332: 1127-1135.

Iegorova O, Fisyunov A, Krishtal O (2010). G-protein-independent modulation of P-type calcium channels by μ-opioids in Purkinje neurons of rat. Neurosci Lett 480: 106-111.

Ikeda K, Kobayashi T, Kumanishi T, Niki H, Yano R (2000). Involvement of G-protein- activated inwardly rectifying K (GIRK) channels in opioid-induced analgesia. Neurosci

Res 38: 113-116.

Ingram SL, Traynor JR (2009). Role of protein kinase C in functional selectivity for desensitisation at the μ-opioid receptor: from pharmacological curiosity to therapeutic potential. Br J Pharmacol 158: 154-156.

294 Ingram S, Wilding TJ, McCleskey EW, Williams JT (1997). Efficacy and kinetics of opioid action on acutely dissociated neurons. Mol Pharmacol 52: 136-143.

Ingram SL, Williams JT (1994). Opioid inhibition of Ih via adenylyl cyclase. Neuron 13:

179-186.

James JR, Oliveira MI, Carmo AM, Iaboni A, Davis SJ (2006). A rigorous experimental framework for detecting protein oligomerisation using bioluminescence resonance energy transfer. Nat Methods 3: 1001-1006.

Johnson EA, Oldfield S, Braksator E, Gonzalez-Cuello A, Couch D, Hall KJ (2006).

Agonist-selective mechanisms of mu-opioid receptor desensitisation in human embryonic kidney 293 cells. Mol Pharmacol 70: 676-685.

Jordan B, Devi LA (1998). Molecular mechanisms of opioid receptor signal transduction.

Brit J Anaesth 81: 12-19.

Juhasz JR, Hasbi A, Rashid AJ, So CH, George SR, O’Dowd BF (2008). Mu-opioid receptor heterooligomer formation with the dopamine D1 receptor as directly visualized in living cells. Eur J Pharmacol 581: 235-243.

Kam AY, Chan AS, Wong YH (2004). Phosphatidylinositol-3 kinase is distinctively required for mu-, but not kappa-opioid receptor-induced activation of c-Jun N-terminal kinase. J Neurochem 89: 391-402.

295 Kapur S, Sharad S, Singh RA, Gupta AK (2007). A118G polymorphism in mu opioid receptor gene (oprm1): association with opiate addiction in subjects of Indian origin. J

Integr Neurosci 6: 511-522.

Karlin A (1967). On the application of “a plausible model” of allosteric proteins to the receptor for acetylcholine. J Theor Biol 16: 306-320.

Kelly E (2013). Efficacy and ligand bias at the μ-opioid receptor. Br J Pharmacol 169:

1430-1446.

Kelly E, Bailey CP, Henderson G (2008). Agonist-selective mechanisms of GPCR desensitisation. Br J Pharmacol 153 Suppl 1: S379-398.

Kenakin T (2002). Efficacy at G-protein-coupled receptors. Nat Rev Drug Discov 1: 103-

110.

Kenakin T, Christopoulos A (2013). Signalling bias in new drug discovery: detection, quantification and therapeutic impact. Nat Rev Drug Discov 12: 205-216.

Kenakin T, Miller LJ (2010). Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery.

Pharmacol Rev 62: 265-304/

Kieffer, BL (1995). Recent advances in molecular recognition and signal transduction of active peptides: Receptors for opioid peptides. Cell Mol Neurobiol 15: 615-635.

296 Kieffer BL (1999). Opioids: first lessons from knockout mice. Trends Pharmacol Sci 20:

19-26.

Kieffer BL, Befort K, Gaveriaux-Ruff C, Hirth CG (1992). The delta-opioid receptor: isolation of a cDNA by expression cloning and pharmacological characterisation. Proc

Natl Acad Sci USA 89: 12048-12052.

Kieffer BL, Gaveriaux-Ruff C (2002). Exploring the opioid system by gene knockout.

Prog Neurobiol 66: 285-306.

Kim CJ, Rhee JS, Akaike N (1997). Modulation of high-voltage activated Ca2+ channels in the rat periaqueductal gray neurons by mu-type opioid agonist. J Neurophysiol 77:

1418-1424.

Kim KJ, Kim HE, Lee KH, Han W, Yi MJ, Jeong J et al (2004). Two-promoter vector is highly efficicent for overproduction of protein complexes. Protein Sci 13: 1698-1703.

Kim KS, Lee KW, Lee KW, Im JY, Yoo JY, Kim SW et al. (2006). Adenylyl cyclase type

5 (AC5) is an essential mediator of morphine action. Proc Natl Acad Sci USA 103: 3908-

3913.

Kim SG, Kim CM, Kang DH, Kim YJ, Byun WT, Kim SY et al. (2004). Association of functional opioid receptor genotypes with alcohol dependence in Koreans. Alcohol Clin

Exp Res 28: 986-990.

297 Klepstad P, Rakvag TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC et al. (2004). The

118A > G polymorphism in the human μ-opioid receptor gene may increase morphine requirements in patients with pain caused by malignant disease. Acta Anaesthesiol Scand

48: 1232-1239.

Knapman A, Connor M (2014). Cellular signalling of non-synonymous single nucleotide polymorphisms of the human µ-Opioid receptor (OPRM1). Br J Pharmacol, In Press doi:

10.1111/bph.12644.

Kovacs KM, Szakall I, O’Brien D, Wang R, Vinod KY Saito M et al. (2005). Decreased oral self-administration of alcohol in kappa-opioid receptor knock-out mice. Alcohol Clin

Exp Res 29: 730-738.

Krillov L, Nguyen A, Miyazaki T, Unson CG, Williams R, Lee NH et al. (2011). Dual mode of internalisation: role of PKCα, GRKs and β-arrestins. Exp Cell

Res 317: 2981-2994.

Kristiansen K (2004). Molecular mechanisms of ligand binding, signalling, and regulation within the superfamily of G-protein-coupled receptors: molecular modeling and mutagenesis approaches to receptor structure and function. Pharmacol Ther 103: 21-80.

Kuszak AJ, Pitchiaya S, Anand JP, Mosberg HI, Walter NG, Sunahara RK (2009).

Purification and functional reconstitution of monomeric μ-opioid receptors: allosteric modulation of agonist binding by Gi2. J Biol Chem 284: 26732-26741.

298 Land BB, Bruchas MR, Lemos JC, Xu M, Melief EJ, Chavkin C (2008). The dysphoric component of stress is encoded by activation of the dynorphin-kappa opioid system. J

Neurosci 28: 407-414.

Langelaan DN, Reddy T, Banks AW, Dellaire G, Dupre DJ, Rainey JK (2013). Structural features of the N-terminal tail and first transmembrane segment implicated in ligand binding and receptor trafficking. Biochim Biophys Acta 1828: 1471-1483.

Laugwitz KL, Offermanns S, Spicher K, Schultz G (1993). µ and δ opioid receptors differentially couple to G protein subtypes in membranes of human neuroblastoma SH-

SY5Y cells. Neuron 10: 233-242.

Laurila JM, Wissel G, Xhaard H, Ruuskanen JO, Johnson MS, Scheinin M (2011).

Involvement of the first transmembrane segment of human α2-adrenoceptors in the subtype-selective binding of chlorpromazine, spiperone and spiroxatrine. Br J Pharmacol

164: 1558-1572.

Law PY, Wong YH, Loh HH (2000). Molecular mechanisms and regulation of opioid receptor signalling. Annu Rev Pharmacol Toxicol 40: 389-430.

Le Merrer J, Rezai X, Scherrer G, Becker JA, Kieffer BL (2013). Impaired - dependent and facilitated striatum-dependent behaviours in mice lacking the δ opioid receptor. Neuropsychopharmacology 38: 1050-1059.

Lefkowitz RJ (1998). G protein couple receptors III. New roles for receptor kinases and β- arrestins in receptor signalling and desensitisation. J Biol Chem 273: 18677-18680.

299

Levitt ES, Williams JT (2012). Morphine desensitisation and cellular tolerance are distinguished in rat locus ceruleus neurons. Mol Pharmacol 82: 983-992.

Li CH, Chung D (1976). Isolation and structure of an untriakontapeptide with opiate activity from camel pituitary glands. Proc Natl Acad Sci U S A 73: 1145-1148.

Li S, Zhu J, Chen C, Chen YW, Deriel JK, Ashby B et al. (1993). Molecular cloning and expression of a rat kappa opioid receptor. Biochem J 295: 629-633.

Liu X, Constantinescu SN, Sun Y, Bogan JS, Hirsch D, Weinberg RA et al (2000).

Generation of mammalian cells stably expressing multiple genes at predetermined levels.

Anal Biochem 280: 20-28.

Liu-Chen LY (2004). Agonist-induced regulation and trafficking of κ opioid receptors. Life

Sciences 75: 511-536.

Loh el W, Fann CS, Chang YT, Chang CJ, Cheng AT (2004). Endogenous opioid receptor genes and alcohol dependence among Taiwanese Han. Alcohol Clin Exp Res 28: 15-19.

Lord JAH, Waterfield AA, Hughes J, Kosterlitz HW (1977). Endogenous opioid peptides: multiple agonists and receptors. Nature 267: 495-499.

Lotsch J, Geisslinger G (2005). Are μ-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol Med 11: 82-89.

300 Lotsch J, Skarke C, Grosch S, Darimont J, Schmidt H, Geisslinger G (2002). The polymorphism A118G of the human μ-opioid receptor gene decreases the pupil constrictory effect of morphine-6-glucuronide but not that of morphine. Pharmacogenetics

12: 3-9.

Luttrell LM (2002). Activation and targeting of mitogen-activated protein-kinases by G- protein-coupled receptors. Can J Physiol Pharmacol 80: 375-382.

Lutz PE, Kieffer BL (2013). Opioid receptors: distinct roles in mood disorders. Trends

Neurosci 36: 195-206.

Mabbott V, Robinson M, Segrave A, Brennan Q (2010). Australian Statistics on Medicines

2010. Accessed online 20/4/14 at http://www.pbs.gov.au/statistics/asm/2010/australian- statistics-on-medicine-2010.pdf.

Macey TA, Lowe JD, Chavkin C (2006). Mu opioid receptor activation of ERK1/2 is

GRK3 and arrestin dependent in striatal neurons. J Biol Chem 281: 34515-34524.

Mague SD, Blendy JA (2010). OPRM1 SNP (A118G): Involvement in disease development, treatment response, and animal models. Drug Alcohol Depend 108: 172-182.

Mague SD, Isiegas C, Huang P, Liu-Chen LY, Lerman C, Blendy JA (2009). Mouse model of OPRM1 (A118G) polymorphism has sex-specific effects on drug-mediated behavior.

Proc Natl Acad Sci USA 106: 10847-10852.

301 Manglik A, Kruse AC, Kobilka TS, Thian FS, Mathiesen JM, Sunahara RK et al. (2012).

Crystal structure of the µ-opioid receptor bound to a morphinan antagonist. Nature 485:

321-326.

Mansour A, Fox CA, Akil H, Watson SJ (1995). Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications. Trends Neurosci 18: 22-29.

Mansour A, Fox CA, Thompson RC, Akil H, Watson SJ (1994). Mu-opioid receptor mRNA expression in the rat CNS: comparison to mu-receptor binding. Brain Res 643:

245-265.

Marker CL, Cintora SC, Roman MI, Stoffel M, Wickman K (2002). Hyperalgesia and blunted morphine analgesia in G protein-gated potassium channel subunit knockout mice.

Neuroreport 13: 2509-2513.

Marker CL, Lujan R, Loh HH, Wickman K (2005). Spinal G-protein-gated potassium channels contribute in a dose-dependent manner to the analgesic effect of mu- and delta- but not kappa-opioids. J Neurosci 25: 3551-3559.

Marker CL, Stoffel M, Wickman K (2004). Spinal G-protein-gated K+ channels formed by

GIRK1 and GIRK2 subunits modulate thermal nociception and contribute to morphine analgesia. J Neurosci 24: 2806-2812.

Martin M, Matifas A, Maldonado R, Kieffer BL (2003). Acute antinociceptive responses in single and combinatorial opioid receptor knockout mice: distinct mu, delta and kappa tones. Eur J Neurosci 17: 701-708.

302 Martin WR, Eades CG, Thompson JA, Huppler RE, Gilbert PE (1976). The effects of morphine- and -like drugs in the nondependent and morphine-dependent chronic spinal dog. J Pharmacol Exp Ther 197: 517-532.

Marquez P, Borse J, Nguyen AT, Hamid A, Lufty K (2008). The role of the opioid receptor-like (ORL1) receptor in motor stimulatory and rewarding actions of buprenorphine and morphine. Neuroscience 155: 597-602.

Marquez P, Nguyen AT, Hamid A, Lutfy K (2008). The endogenous OFQ/N/ORL-1 receptor system regulates the rewarding effects of cocaine. Neuropharmacology 54: 564-

568.

Massotte D, Brillet K, Kieffer BL, Milligan G (2002). Agonists activate Gi1α or Gi2α fused to the human mu opioid receptor differently. J Neurochem 81: 1372-82.

Matthes HWD, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I et al. (1996).

Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the µ-opioid-receptor gene. Nature 383: 819-823.

Matthes HW, Smadja C, Valverde O, Vonesch JL, Foutz AS, Boudinot E et al.(1998).

Activity of the delta-opioid receptor is partially reduced, whereas activity of the kappa- receptor is maintained in mice lacking the mu-receptor. J Neurosci 18: 7285-7295.

Mazarakou G, Georgoussi Z (2005). STAT5A interacts with and is phosphorylated upon activation of the mu-opioid receptor. J Neurochem 93: 918-931.

303 McCudden CR, Hains MD, Kimple RJ, Siderovski DP, Willard FS (2005). G-protein signalling: back to the future. Cell Mol Life Sci 62: 551-577.

McLaughlin JP, Marton-Popovici M, Chavkin C (2003). κ Opioid receptor antagonism and prodynorphin gene disruption block stress-induced behavioural responses. J Neurosci 23:

5674-5683.

McPherson J, Rivero G, Baptist M, LLorente J, Al-Sabah S, Krasel C et al. (2010). μ-

Opioid receptors: correlation of agonist efficacy for signalling with ability to activate internalisation. Mol Pharmacol 78: 756-766.

Meis S (2003). Nociceptin/orphanin FQ: actions within the brain. Neuroscientist 9: 158-

168.

Melief EJ, Miyatake M, Bruchas MR, Chavkin C (2010). Ligand-directed c-Jun N-terminal kinase activation disrupts opioid receptor signalling. Proc Natl Acad Sci U S A 107:

11608-11613.

Meng F, Xie GX, Thompson RC, Mansour A, Goldstein A, Watson SJ et al. (1993).

Cloning and pharmacological characterisation of a rat kappa opioid receptor. Proc Natl

Acad Sci USA 90: 9954-9958.

Meunier J-C (1997). Nociceptin/orphanin FQ and the opioid receptor-like ORL1 receptor.

Eur J Pharmacol 340: 1-15.

304 Meunier J-C, Mollereau C, Suaudeau C, Moisand C, Alvinerie P, Butour JL et al. (1995).

Isolation and structure of the endogenous agonist of opioid receptor-like ORL-1 receptor.

Nature 377: 532-535.

Milligan G (2009). G protein-coupled receptor hetero-dimerisation: contribution to pharmacology and function. Br J Pharmacol 158: 5-14.

Miranda R, Ray L, Justus A, Meyerson LA, Knopik VS, McGeary J et al. (2010). Initial evidence of an association between OPRM1 and adolescent alcohol misuse. Alcohol Clin

Exp Res 34: 112-122.

Mitrovic I, Margeta-Mitrovic M, Bader S, Stoffel M, Jan LY, Basbaum AI (2003).

Contribution of GIRK2-mediated postsynaptic signalling to opiate and alpha 2-adrenergic analgesia and analgesic sex differences. Proc Natl, Acad Sci USA 100: 271-276.

Mokrosinski J, Frimurer TM, Siversten B, Schwartz TW, Holst B (2012). Modulation of constitutive activity and signalling bias of the ghrelin receptor by conformational constraint in the second extracellular loop. J Biol Chem 287: 33488-33502.

Mollereau C, Parmentier M, Mailleux P, Butour JL, Moisand C, Chalon P et al. (1994).

ORL1, a novel member of the opioid receptor family. Cloning, functional expression and localisation. FEBS Lett 341: 33-38.

Moore RA, McQuay HJ (2005). Prevalence of opioid adverse events in chronic non- malignant pain: systematic review of randomized trials of oral opioids. Arthrit Res Ther 7:

R1046-R1051.

305

Morikawa H, Fukuda K, Kato S, Mori K, Higashida H (1995). Coupling of the cloned μ-

Opioid receptor with the ω-conotoxin-sensitive Ca2+ current in NG108-15 cells. J

Neurochem 65: 1403-1406.

Mouledous L, Diaz MF, Gutstein HB (2004). Modulation of extracellular signal-regulated kinase (ERK) activity by acute and chronic opioid treatment in neuronal and clial cell lines. J Neurochem 90: 1371-1377.

Mura E, Govoni S, Racchi M, Carossa V, Ranzani GN, Allegri M, van Schaik RHN

(2013). Consequences of the 118A>G polymorphism in the OPRM1 gene: translation from bench to bedside? J Pain Res 6: 331-53.

Nadal X, Banos JE, Kieffer BL, Maldonado R (2006). Neuropathic pain is enhanced in delta-opioid receptor knockout mice. Eur J Neurosci 23: 830-834.

Narita M, Ohsawa M, Mizoguchi H, Aoki T, Suzuki T, Tseng LF (2000). Role of the phosphatidylinositol-specific phospholipase C pathway in delta-opioid receptor-mediated antinociception in the mouse spinal cord. Neuroscience 99: 327-331.

Neilan CL, Husbands SM, Breedan S, Ko MC, Aceto MD, Lewis JW et al. (2004).

Characterisation of the complex morphinan derivative BU72 as a high efficacy, long- lasting μ-opioid receptor agonist. Eur J Pharmacol 499: 107-116.

Nevo I, Avidor-Reiss T, Levy R, Bayewitch M, Vogel Z (2000). Acute and chronic activation of the µ-opioid receptor with the endogenous ligand endomorphin differentially regulates adenylyl cyclase isozymes. Neuropharmacology 39: 364-371.

306

Nickolls SA, Waterfield A, Williams RE, Kinloch RA (2011). Understanding the effect of different assay formats on agonist parameters: a study using the μ-opioid receptor. J

Biomol Screen 16: 706-716.

Nieto MM, Guen SL, Kieffer BL, Rogues BP, Noble F (2005). Physiological control of emotion-related behaviours by endogenous enkephalins involves essentially the delta opioid receptors. Neuroscience 135: 305-313.

Nishizawa D, Han W, Hasegawa J, Ishida T, Numata Y, Sato T et al. (2006). Association of μ-opioid receptor gene polymorphism A118G with alcohol dependence in a Japanese population. Neuropsychobiology 53: 137-141.

Noble M, Treadwell JR, Tregear SJ, Coates VH, Wiffen PJ, Akafomo C et al. (2010).

Long-term opioid management for chronic noncancer pain. Cochrane Database of

Systemic Reviews 1: CD006605.

Ocana M, Cendan CM, Cobos EJ, Entrena JM, Baeyens JM (2004). Potassium channels and pain: present realities and future opportunities. Eur J Pharmacol 500: 203-219.

Oertel BG, Kettner M, Scholich K, Renne C, Roskam B, Geisslinger G et al. (2009). A common human μ-opioid receptor genetic variant diminishes the receptor signalling efficacy in brain regions processing the sensory information of pain. J Biol Chem 284:

6530-6535.

Okajima F, Tomura H, Kondo Y (1993). Enkephalin activates the phospholipase C/Ca2+ system through cross-talk between opioid receptors and P2-purinergic or

307 receptors in NG 108-15 cells. A permissive role for pertussis toxin-sensitive G proteins.

Biochem J 290 (Pt 1): 241-247.

Osborne PB, Chieng B, Christie MJ (2000). Morphine-6β-glucuronide has a higher efficacy than morphine as a μ-opioid receptor agonist in the rat locus coeruleus. Br J

Pharmacol 131: 1422-1428.

Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, Fox BA et al. (2000).

Crystal structure of rhodopsin: A G protein-coupled receptor. Science 289: 739-745.

Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP, Buneman OP et al.

(2014). The IUPHAR/BPS Guide to PHARMACOLOGY: an expert-driven knowledgebase of dug targets and their ligands. Nucleic Acids Res: 42 D1098-1106.

Payne AJ, Gerdes BC, Kaja S, Koulen P (2013). Insert sequence length determines transfection efficiency and gene expression levels in bicistronic mammalian expression vectors. Int J Biochem Mol Biol 4: 201-208.

Peckys D, Landwehrmeyer GB (1999). Expression of mu, kappa, and delta opioid receptor messenger RNA in the human CNS: A 33P in situ hybridisation study. Neuroscience 88:

1093-1135.

Pepper CM, Henderson G (1980). Opiates and opioid peptides hyperpolarize locus coeruleus neurons in vitro. Science 209: 394-395.

308 Pert CB, Snyder SH (1973). Opiate receptor: demonstration in nervous tissue. Science 179:

1011-1014.

Pfutzner W (2008). Retroviral bicistronic vectors. Drug News Perspect 21: 473-480.

Piechowski CL, Rediger A, Lagemann C, Muhlhaus J, Muller A, Pratzka J et al (2013).

Inhibition of melanocortin-4-receptor dimerisation by substitutions in intracellular loop 2.

J Mol Endocrinol 51: 109-118.

Pineyro G, Archer-Lahlou E (2007). Ligand-specific receptor states: implications for opiate receptor signalling and regulation. Cell Signal 19: 8-19.

Polakiewicz RD, Schieferi SM, Dorner LF, Kansara V, Comb MJ (1998). A mitogen- activated protein-kinase pathway is required for mu-opioid receptor desensitisation. J Biol

Chem 273: 12402-12406.

Pradhan A, Smith M, McGuire B, Evans C, Walwyn W (2013). Chronic inflammatory injury results in increased coupling of delta opioid receptors to voltage-gated Ca2+ channels. Mol Pain 9: PMC3621800.

Raehal KM, Bohn LM (2011). The role of β-arrestin 2 in the severity of antinociceptive tolerance and physical dependent induced by different opioid pain therapeutics.

Neuropharmacology 60: 58-65.

Raehal KM, Walker JK, Bohn LM (2005). Morphine side effects in β-arrestin 2 knockout mice. J Pharmacol Exp Ther 314: 1195-1201.

309 Raffa RB, Martinez RP, Connelly CD (1994). G-protein antisense oligodeoxyribonucleotides and μ-opioid supraspinal antinociception. Eur J Pharmacol

258: R5-7.

Rajagopal S, Bassoni DL, Campbell JJ, Gerard NP, Gerard C, Wehrman TS (2013). Biased agonism as a mechanism for differential signalling by chemokine receptors. J Biol Chem

288: 35039-35048.

Ramchandani VA, Umhau J, Pavon FJ, Ruiz-Velasco V, Margas W, Sun H et al. (2011). A genetic determinant of the striatal dopamine response to alcohol in men. Mol Psychiatry

16: 809-817.

Rana BK, Shilina T, Insel PA (2001). Genetic variations and polymorphisms of G protein- coupled receptors: functional and therapeutic implications. Annu Rev Pharmacol Toxicol

41: 593-624.

Randall AD (1998). The molecular basis of voltage-gated Ca2+ channel diversity: is it time for T? J Membr Biol 161: 207-213.

Raote I, Bhattacharyya S, Panicker MM (2013). Functional selectivity in serotonin receptor 2A (5-HT2A) endocytosis, recycling, and phosphorylation. Mol Pharmacol 83: 42-

50.

Ravindranathan A, Joslyn G, Robertson M, Schuckit MA, Whistler JL, White RL (2009).

Functional characterisation of human variants of the mu-opioid receptor gene. Proc Natl

Acad Sci USA 106: 10811-10816.

310

Ray LA, Hutchison KE (2004). A polymorphism of the μ-opioid receptor gene (OPRM1) and sensitivity to the effects of alcohol in humans. Alcohol Clin Exp Res 28: 1789-1795.

Ray LA, Hutchison KE (2007). Effects of naltrexone on alcohol sensitivity and genetic moderators of medication response: a double-blind placebo-controlled study. Arch Gen

Psychiatry 64: 1069-1077.

Ray LA, Miranda R Jr, Tidey JW, McGeary JE, MacKillop J, Gwaltney CJ et al. (2010).

Polymorphisms of the μ-opioid receptor and dopamine D4 receptor genes and subjective responses to alcohol in the natural environment. J Abnorm Psychol 119: 115-125.

Reinscheid RK, Nothacker H-P, Civelli O (2000). The orphanin FQ/nociceptin gene: structure, tissue distribution of expression and functional implications obtained from knockout mice. Peptides 21: 901-906.

Reyes-Gibby CC, Shete S, Rakvag T, Bhat SV, Skorpen F, Bruera E et al. (2007).

Exploring joint effects of genes and the clinical efficacy of morphine for cancer pain:

OPRM1 and COMT gene. Pain 130: 25-30.

Rivero G, LLorente J, McPherson J, Cooke A, Mundell SJ, McArdle CA et al (2012).

Endomorphin-2: a biased agonist at the μ-opioid receptor. Mol Pharmacol 82: 178-188.

Rizzi A, Molinari S, Marti M, Marzola G, Calo’ G (2011). Nociceptin/orphanin FQ receptor knockout rats: in vitro and in vivo studies. Neuropharmacology 60: 572-579.

311 Roberts AJ, Gold LH, Polis I, McDonald JS, Filliol D, Kieffer BL et al. (2001). Increased ethanol self-administration in delta-opioid receptor knockout mice. Alcohol Clin Exp Res

25: 1249-1256.

Rodrigues AR, Castro MS, Francischi JN, Perez AC, Duarte ID (2005). Participation of

ATP-sensitive K+ channels in the peripheral antinociceptove effect of fentanyl in rats.

Braz J Med Biol Res 38: 91-97.

Romberg RR, Olofsen E, Bijl H, Taschner PE, Teppema LJ, Sarton EY et al. (2005).

Polymorphism of μ-opioid receptor gene (OPRM1:c.118A>G) does not protect against opioid-induced respiratory depression despite reduced analgesic response. Anesthesiology

102: 522-530.

Rommelspacher H, Smolka M, Schmidt LG, Samochowiec J, Hoehe MR (2001). Genetic analysis of the mu-opioid receptor in alcohol-dependent individuals. Alcohol 24: 129-135.

Rosenbaum DM, Rasmussen SG, Kobilka BK (2009). The structure and function of G- protein-coupled receptors. Nature 459: 356-363.

Rothman RB, Bykov V, Xue BG, Xu H, De Costa BR, Jacobson AE et al. (1992).

Interaction of opioid peptides and other drugs with multiple kappa receptors in rat and human brain. Evidence for species differences. Peptides 13: 977-987.

Rusin KI, Moises HC (1995). Mu-Opioid receptor activation reduces multiple components of high-threshold calcium current in rat sensory neurons. J Neurosci 15: 4315-4327.

312 Rutten K, De Vry J, Bruckmann W, Tzschentke TM (2011). Pharmacological blockade or genetic knockout of the NOP receptor potentiates the rewarding effect of morphine in rats.

Drug Alcohol Depend 114: 253-256.

Sadana R, Dessauer CW (2009). Physiological roles for G-protein-regulated adenylyl cyclase isoforms: insights from knockout and overexpression studies. Neurosignals 17: 5-

22.

Saidak Z, Blake-Palmer K, Hay DL, Northup JK, Glass M (2006). Differential activation of G-proteins by μ-opioid receptor agonists. Brit J Pharmacol 147: 671-680.

Sakoori K, Murphy NP (2008). Endogenous nociception (orphanin FQ) suppresses basal hedonic state and acute reward responses to and ethanol, but facilitates chronic responses. Neuropsychopharmacology 33: 877-891.

Salamon Z, Hruby VJ, Tollin G, Cowell S (2002). Binding of agonists, antagonists and inverse agonists to the human δ-opioid receptor produces distinctly different conformational states distinguishable by plasmon-waveguide resonance spectroscopy. J

Pept Res 60: 322-328.

Samama P, Cotecchia S, Costa T, Lefkowitz RJ (1993). A mutation-induced activated state of the β2-adrenergic receptor. Extending the ternary complex model. J Biol Chem 268:

4625-4636.

Samways DS, Henderson G (2006). Opioid elevation of intracellular free calcium: possible mechanisms and physiological relevance. Cell Signal 18: 151-161.

313

Sanchez JA, Gonoi T, Inagaki N, Katada T, Seino S (1998). Modulation of reconstituted

ATP-sensitive K(+) channels by GTP binding proteins in a mammalian cell line. J Physiol

507: 315-24.

Sanchez-Blazquez P, DeAntoio I, Rodriguez-Diaz M, Garzon J (1999). Antisense oligodeoxynucleotide targeting distinct exons of the cloned μ-opioid receptor distinguish between endomorphin-1 and morphine supraspinal antinociception in mice. Antisense

Nucleic Acid Drug Dev 9: 253-260.

Sanchez-Blazquez P, Garcia-Espana A, Garzon J (1995). In vivo injection of antisense oligodeoxynucleotides to Gα subunits and supraspinal analgesia evoked by μ and δ opioid agonists. J Pharmacol Exp Ther 275: 1590-1596.

Sanchez-Blazquez P, Gomez-Serranillos P, Garzon J (2001). Agonists determine the pattern of G-protein activation in μ-opioid receptor-mediated supraspinal analgesia. Brain

Res Bull 54: 229-235.

Sanchez-Blazquez P, Juarros JL, Martinez-Pena Y, Castro MA, Garzon J (1993). Gx/z and

Gi2 transducer proteins on μ/δ opioid-mediated supraspinal antinociception. Life Sci 53:

PL381-386.

Sarne Y, Rubovitch V, Fields A, Gafni M (1998). Dissociation between the inhibitory and stimulatory effects of opioid peptides on cAMP formation in SK-N-SH neuroblastoma cells. Biochem Biophys Res Commun 246: 128-131.

314 Satoh M, Minami M (1995). Molecular pharmacology of the opioid receptors. Pharmacol

Ther 68: 343-364.

Schallmack E, Steiner D, Vogel Z (2006). Adenylyl cyclase type II activity is regulated by two different mechanisms: implications for acute and chronic morphine exposure.

Neuropharmacology 50: 998-1005.

Scherrer G, Befort K, Contet C, Becker J, Matifas A, Kieffer BL (2004). The delta agonists

DPDPE and deltorphin II recruit predominantly mu receptors to produce thermal analgesia: a parallel study of mu, delta and combinatorial opioid receptor knockout mice. Eur J

Neurosci 19: 2239-2248.

Schroeder JE, Fischbach PS, Zheng D, McCleskey EW (1991). Activation of μ-opioid receptors inhibits transient high- and low-threshold Ca2+ currents, but spares a sustained current. Neuron 6: 13-20.

Schroeder JE, McCleskey EW (1993). Inhibition of Ca2+ currents by a μ-opioid in a defined subset of rat sensory neurons. J Neurosci 13: 867-873.

Serohijos AWR, Yin S, Ding F, Gauthier J, Gibson DG, Maixner W et al. (2011).

Structural bias for mu-opioid receptor binding and activation. Structure 19: 1683-1690.

Setiawan E, Pihl RO, Cox SM, Gianoulakis C, Palmour RM, Benkelfat C, Leyton M

(2011). The effect of naltrexone on alcohol’s stimulant properties and self-administration

315 behaviour in social drinkers: influence of gender and genotype. Alcohol Clin Exp Res 35:

1134-1141.

Seward E, Hammond C, Henderson G (1991). Mu-opioid receptor-mediated inhibition of the N-type calcium channel current. Proc Biol Sci 244: 129-135.

Sharif NA, Hughes J (1989). Discrete mapping of brain mu and delta opioid receptors using selective peptides: quantitative autoradiography, species differences and comparison with kappa receptors. Peptides 10: 499-522.

Shenoy SK, Lefkowitz RJ (2005). Seven-transmembrane receptor signalling through β- arrestin. Sci STKE 2005(308): cm10

Shi J, Hui L, Xu Y, Wang F, Huang W, Hu G (2002). Sequence variations in the μ-opioid receptor gene (OPRM1) associated with human addiction to heroin. Hum Mutat 19: 459-

460.

Sia AT, Lim Y, Lim EC, Goh RW, Law HY, Landau R et al (2008). A118G single nucleotide polymorphism of human μ-opioid receptor gene influences pain perception and patient-controlled intravenous morphine consumption after intrathecal morphine for postcesarean analgesia. Anesthesiology 109: 520-526.

Simon EJ, Hiller JM, Edelman I (1973). Stereospecific binding of the potent analgesic [3H]Etorphine to rat-brain homogenate. Proc Natl Acad Sci USA 70: 1947-1949.

316 Simonin F, Valverde O, Smadja C, Slowe S, Kitchen I, Dierich A et al. (1998). Disruption of the κ-opioid receptor gene in mice enhances sensitivity to chemical visceral pain, impairs pharmacological actions of the selective κ-agonist U-50,488H and attenuates morphine withdrawal. EMBO J 17: 886-897.

Singh VK, Bajpai K, Biswas S, Haq W, Khan MY, Mathur KB (1997). Molecular biology of opioid receptors. Neuroimmunomodulation 4: 285-297.

Skalhegg BS, Tasken K (1997). Specificity in the cAMP/PKA signalling pathway.

Differential expression, regulation, and subcellular localisation of subunits of PKA. Front

Biosci 2: d331-d342.

Skarke C, Darimont J, Schmidt H, Geisslinger G, Lotsch J (2003). Analgesic effects of morphine and morphine-6-glucuronide in a transcutaneous electrical pain model in healthy volunteers. Clin Pharmacol Ther 73: 107-121.

Skorpen F, Laugsend EA, Klepstad P, Kaasa S (2008). Variable response to opioid treatment: any genetic predictors in sight? Palliat Med 22: 310-327.

Somogyi AA, Barrat DT, Coller JK (2007) Pharmacogenetics of opioids. Clin Pharmacol

Ther 81: 429-444.

Sora I, Li XF, Funada M, Kinsey S, Uhl GR (1999). Visceral chemical nociception in mice lacking mu-opioid receptors: effects of morphine, SNC80 and U-50,488. Eur J Pharmacol

366: R3-5.

317 Sora I, Takahashi N, Funada M, Ujike H, Revay RS Donovan DM et al. (1997). Opiate receptor knockout mice define mu receptor roles in endogenous nociceptive responses and morphine-induced analgesia. Proc Natl Acad Sci 94: 1544-1549.

Snyder SH (1985). Adenosine as a neuromodulator. Annu Rev Neurosci 8: 103-124.

Sternini C (2001). Receptors and transmission in the brain-gut axis: potential for novel therapies. III. Mu-opioid receptors in the enteric nervous system. Am J Physiol

Gastrointest Physiol 281: G8-15.

Stockton SD Jr, Devi LA (2012). Functional relevance of µ-δ opioid receptor heteromerisation: a role in novel signalling and implications for the treatment of addiction disorders: from a symposium on new concepts in mu-opioid pharmacology. Drug Alcohol

Depend 121: 167-172.

Svoboda KR, Lupica CR (1998). Opioid inhibition of hippocampal interneurons via modulation of potassium and hyperpolarisation-activated cation (Ih) currents. J Neurosci

18: 7084-7098.

Szeto CY, Tang NL, Lee DT, Stadlin A (2001). Association between μ-opioid receptor gene polymorphisms and Chinese heroin addicts. Neuroreport 12: 1103-1106.

Tan EC, Lim EC, Teo YY, Lim Y, Law HW, Sia AT (2009). Ethnicity and OPRM variant independently predict pain perception and patient-controlled analgesia usage for post- operative pain. Mol Pain 5:32 doi: 10.1186/1744-8069-5-32.

318 Tan EC, Tan CH, Karupathivan U, Yap EP (2003). Mu opioid receptor gene polymorphisms and heroin dependence in Asian populations. Neuroreport 14: 569-572.

Tan M, Groszer M, Tan AM, Pandya A, Liu X, Xie CW (2003). Phosphoinositide 3-kinase cascade facilitates mu-opioid desensitisation in sensory neurons by altering G-protein- effector interactions. J Neurosci 23: 10292-10301.

Taylor BK (2009). Spinal inhibitory neurotransmission in neuropathic pain. Curr Pain

Headache Rep 13: 208-214.

Tedford HW, Zamponi GW (2006). Direct G protein modulation of Cav2 calcium channels. Pharmacol Rev 58: 837-862.

Terenius L (1973). Stereospecific interaction between narcotic analgesics and a synaptic plasma membrane fraction of rat . Acta Pharmacol Toxicol 32: 317-320.

Thompson AA, Liu W, Chun E, Katritch V, Wu H, Vardy E et al. (2012). Structure of the nociception/orphanin FQ receptor in complex with a peptide mimetic. Nature 485: 395-

399.

Thompson MD, Percy ME, McIntyre WB, Cole DE (2008). G protein-coupled receptors disrupted in human genetic disease. Methods Mol Biol 448: 109-137.

Thompson RC, Mansour A, Akil H, Watson SJ (1993). Cloning and pharmacological characterisation of a rat mu opioid receptor. Neuron 11: 903-913.

319 Thorsell A (2013). The μ-opioid receptor and treatment response to naltrexone. Alcohol

Alcohol 48: 402-408.

Torrecilla M, Marker CL , Cintora SC, Stoffel M, Williams JT, Wickman K (2002). G- protein-gated potassium channels containing Kir3.2 and Kir3.3 subunits mediate the acute inhibitory effects of opioids on locus ceruleus neurons. J Neurosci 22: 4328-4334.

Trapaidze N, Gomes I, Cvejic S, Bansinath M, Devi LA (2000). Opioid receptor endocytosis and activation of MAP kinase pathway. Brain Res Mol Brain Res 76: 220-

228.

Tsai FF, Fan SZ, Yang YM, Chien KL, Su YN, Chen LK (2010). Human opioid μ-receptor

A118G polymorphism may protect against central pruritis by epidural morphine for post- cesarean analgesia. Acta Anaesthesiol Scand 54: 1265-1269.

Van den Wildenberg E, Wiers RW, Dessers J, Janssen RG, Lambrichs EH, Smeets HJ et al. (2007). A functional polymorphism of the μ-opioid receptor gene (OPRM1) influences cue-induced craving for alcohol in male heavy drinkers. Alcohol Clin Exp Res 31: 1-10.

Van Ree JM, Gerrits MA, Vanderschuren LJ (1999). Opioids, reward and addiction: An encounter of biology, psychology, and medicine. Pharmacol Rev 51: 341-396.

Varga EV, Stropova D, Rubenzik M, Wang M, Landsman RS, Roeske WR et al (1998).

Identification of adenylyl cyclase isoenzymes in CHO and B82 cells. Eur J Pharmacol

348: R1-R2.

320 Vassart G, Costagliola S (2011). G protein-coupled receptors: mutations and endocrine diseases. Nat Rev Endocrinol 7: 362-372.

Vaughan CW, Bagley EE, Drew GM, Schuller A, Pintar JE, Hack SP et al. (2003).

Cellular actions of opioids on periaqueductal grey neurons from C57B16/J mice and mutant mice lacking MOR-1. Br J Pharmacol 139: 362-367.

Vaughan CW, Ingram SL, Connor MA, Christie MJ (1997). How opioids inhibit GABA- mediated neurotransmission. Nature 390: 611-614.

Violin JD, Lefkowitz RJ (2007). β-arrestin-biased ligands at seven-transmembrane receptors. Trends Pharmacol Sci 28: 416-422.

Volkow ND, McLellan TA (2011). Curtailing diversion and abuse of opioid analgesics without jeopardizing pain treatment. JAMA 305: 1346 – 1347.

Wada Y, Yamashita T, Imai K, Miura R, Takao K, Nishi M et al. (2000). A region of the sulfonylurea receptor critical for a modulation of ATP-sensitive K+ channels by G-protein

βγ subunits. EMBO J 19: 4915-4925.

Wagley Y, Hwang CK, Lin HY, Kam AF, Law PY, Loh HH et al. (2013). Inhibition of c-

Jun NH2-terminal kinase stimulates μ-opioid receptor expression via p38 MAPK-mediated nuclear NF-κB activation in neuronal and non-neuronal cells. Biochim Biophys Acta 1833:

1476-1488.

321 Walter C, Lotsch J (2009). Meta-analysis of the relevance of the OPRM1 118A>G genetic variant for pain treatment. Pain 146: 270-275.

Wang D, Quillan JM, Winans K, Lucas JL, Sadee W (2001). Single nucleotide polymorphisms in the human µ opioid receptor gene alter basal G protein coupling and calmodulin binding. J Biol Chem 276: 34624-34630.

Wang H, Heijnen CJ, van Velthoven CT, Willemen HL, Ishikawa Y, Zhang X et al.

(2013). Balancing GRK2 and EPAC1 levels prevents and relieves chronic pain. J Clin

Invest 123: 5023-5034.

Wang J, Charboneau R, Barke RA, Loh HH, Roy S (2002). Mu-opioid receptor mediates chronic restraint stress-induced lymphocyte apoptosis. J Immunol 169: 3630-3636.

Wang QP, Zadina JE, Guan JL, Shioda S (2002). Morphological studies of the endomorphinergic neurons in the central nervous system. Jpn J Pharmacol 89: 209-215.

Wang Y, Sun J, Tao Y, Chi Z, Liu J (2010). The role of κ-opioid receptor activation in mediating antinociception and addicition. Acta Pharmacol Sin 31: 1065-1070.

Wang YJ, Huang P, Ung A, Blendy JA, Liu-Chen LY (2012). Reduced expression of the

µ-opioid receptor in some, but not all, brain regions in mice with OPRM1 A112G.

Neuroscience 205: 178-184.

Wee S, Koob GF (2010). The role of the dynorphin-kappa opioid system in the reinforcing effects of drug abuse. Psychopharmacology 210: 121-135.

322

Wei L, Loh, HH (2011). Transciptional and epigenetic regulation of opioid receptor genes

– present and future. Annu Rev Pharmacol Toxicol 51: 75-97.

Werry TD, Wilkinson GF, Willars GB (2003). Mechanisms of cross-talk between G- protein-coupled receptors resulting in enhanced release of intracellular Ca2+. Biochem J

374: 281-296.

Wiers RW, Rinck M, Dictus M, van den Wildenberg E (2009). Relatively strong automatic appetitive action-tendencies in male carriers of the OPRM1 G-allele. Genes Brain Behav

8: 101-106.

Wilbanks AM, Laporte SA, Bohn LM, Barak LS, Caron MG (2002). Apparent loss-of- function mutant GPCRs revealed as constitutively desensitized receptors. Biochemistry 41:

11981-11989.

Williams JT, Ingram SL, Henderson G, Chavkin C, von Zastrow M, Schulz S, et al (2013).

Regulation of µ-opioid receptors: desensitisation, phosphorylation, internalisation, and tolerance. Pharmacol Rev 65: 223-254.

Wimpey TL, Chavkin C (1991). Opioids activate both an inward rectifier and a novel voltage-gated potassium conductance in the hippocampal formation. Neuron 6: 281-289.

Wu H, Wacker D, Mileni M, Katritch V, Han GW, Vardy E et al. (2012). Structure of the human κ-opioid receptor in complex with JDTic. Nature 485: 327-332.

323 Wylot B, Tworus K, Okrasa S (2013). The effects of mu-, delta-, and kappa-opioid receptor activation on luteinizing and follicle-stimulating hormone secretion from porcine pituitary cells. J Physiol Pharmacol 64: 505-511.

Xie W, Samoriski GM, McLaughlin JP, Romoser VA, Smrcka A, Hinkle PM et al. (1999).

Genetic alteration of phospholipase Cβ3 expression modulates behavioural and cellular responses to μ opioids. Proc Natl Acad Sci U S A 96: 10385-10390.

Xu H, Gintzler AR (1992). Opioid enhancement of evoked [Met5]encephalin release requires activation of cholinergic receptors: possible involvement of intracellular calcium.

Proc Natl Acad Sci 89: 1978-1982.

Yasuda K, Raynor K, Kong H, Breder CD, Takeda J, Reisine T et al. (1993). Cloning and functional comparison of kappa and delta opioid receptors from mouse brain. Proc Natl

Acad Sci USA 90: 6736-6740.

Yoshimura H, Ida S, Oguri K, Tsukamoto H (1973). Biochemical basis for analgesic activity of morphine-6-glucuronide-1. Penetration of morphine-6-glucuronide in the brain of rats. Biochem Pharmacol 22: 1423-1430.

Yoshimura M, Ikeda H, Tabakoff B (1996). Mu-opioid receptors inhibit dopamine stimulated activity of type V adenylyl cyclase but enhance dopamine-stimulated activity of type VII adenylyl cyclase. Mol Pharmacol 50: 43-51.

Zadina JE, Hackler L, Ge LJ, Kastin AJ (1997). A potent and selective endogenous agonist for the μ-opiate receptor. Nature 386: 499-502.

324

Zaki PA, Keith DE Jr, Brine GA, Carroll FI, Evans CJ (2000). Ligand-induced changes in surface μ-opioid receptor number: relationship to G protein activation. J Pharmacol Exp

Ther 292: 1127-1134.

Zarrindast MR, Jafari MR, Ahmadi S, Djahanguiri B (2004). Influence of central administration ATP-dependent K+ channel on morphine state-dependent memory of passive avoidance. Eur J Pharmacol 487: 143-148.

Zarrindast MR, Sattari-Naeini M, Khalizadeh A (2007). Involvment of glucose and ATP- sensitive potassium (K+) channels on morphine-induced conditioned place preference. Eur

J Pharmacol 573: 133-138.

Zhang J, Ferguson SS, Barak LS, Bodduluri SR, Laporte SA Law PY et al. (1998). Role for G protein-coupled receptor kinase in agonist-specific regulation of μ-opioid receptor responsiveness. Proc Natl Acad Sci U S A 95: 7157-7162.

Zhang F, Steinberg SF (2013). S49G and R389G polymorphisms of the β-adrenergic receptor influence signalling via the cAMP-PKA and ERK pathways. Physiol Genomics

45: 1186-1192.

Zhang H, Cuevas J (2005). σ-receptor activation blocks potassium channels and depresses neuroexcitability in rat intracardiac neurons. J Pharmacol Exp Ther 313: 1387-1396.

Zhang L, Loh HH, Law PY (2013). A novel noncanonical signalling pathway for the µ- opioid receptor. Mol Pharmacol 84: 844-853.

325 Zhang W, Chang YZ, Kan QC, Zhang LR, Lu H, Chu QJ et al. (2010). Association of human μ-opioid receptor gene polymorphism A118G with fentanyl analgesia consumption in Chinese gynaecological patients. Anaesthesia 65: 130-135.

Zhang W, Yuan JJ, Kan QC, Zhang LR, Chang YZ, Wang ZY (2011). Study of the

OPRM1 A118G genetic polymorphism associated with postoperative nausea and induced by fentanyl intravenous analgesia. Minerva Anestesiol 77: 33-39.

Zhang Y, Wang D, Johnson AD, Papp AC, Sadee W (2005). Allelic expression imbalance of human mu opioid receptor (OPRM1) caused by variant A118G. J Biol Chem 280:

32618-24.

Zheng H, Loh HH, Law PY (2008). Β-arrestin-dependent µ-opioid receptor-activated extracellular signal-regulated kinases (ERKs) translocate to nucleus in contrast to G protein-dependent ERK activation. Mol Pharmacol 73: 178-190.

Zheng H, Pearsall EA, Hurst DP, Zhang Y, Chu J, Zhou Y et al (2012). Palmitoylation and membrane cholesterol stabilize μ-opioid receptor homodimerisation and G protein coupling. BMC Cell Biol 13: 6.

Zhu Y, King MA, Schuller AG, Nitsche JE, Reidl M, Elde RP et al. (1999). Retention of supraspinal delta-like analgesia and loss of morphine tolerance in delta opioid receptor knockout mice. Neuron 24: 243-252.

Zimprich A, Simon T, Hollt V (1995). Transfected rat mu opioid receptors (rMOR1 and rMOR1B) stimulate phospholipase C and Ca2+ mobilisation. Neuroreport 7: 54-56.

326

327

328 Appendix

Fluorescence-Based, High-Throughput Assays for

µ-Opioid Receptor Activation using Membrane Potential-Sensitive Dye

Alisa Knapman and Mark Connor

This book chapter was accepted for publication in the book: “Opioid Receptors: Methods and Protocols”, edited by Dr Santi Spampinato. This volume will be published in 2014 as part of the series Methods in Molecular Biology by Springer Protocols Mark Connor assisted with preparation of the manuscript.

329 Summary

The development of new and improved opioid analgesics requires high throughput screening (HTS) methods to identify potential therapeutics from large libraries of lead compounds. Here we describe two simple, real-time fluorescence-based assays of μ-opioid receptor activation that may be scaled up for HTS. In AtT-20 cells expressing the µ-opioid receptor (MOPr), opioids activate endogenous G protein gated inwardly rectifying K channels (GIRK channels), leading to membrane hyperpolarisation. In Chinese hamster ovary cells expressing MOPr, adenylyl cyclase activation via forskolin results in membrane hyperpolarisation, which is inhibited by opioids. Changes in membrane potential can be measured using a proprietary membrane potential-sensitive dye. In contrast to many HTS methods currently available, these assays reflect naturalistic coupling of the receptor to effector molecules.

Keywords

GIRK, adenylyl cyclase, µ-opioid receptor, membrane potential, high-throughput screening, AtT-20, CHO, fluorescent assay.

330 1. Introduction

The µ-opioid receptor (MOPr) is a G protein coupled receptor (GPCR), and is the primary target for opioid analgesics (1). There is increasing evidence that different ligands can stabilise GPCRs including MOPr in various active conformations, leading to the preferential activation of distinct signalling pathways via Gα and Gβγ subunits (2 – 3). The development of novel opioid pharmacotherapies is now beginning to focus on identifying

MOPr ligands that can selectively activate a subset of effector pathways associated with analgesia, without activating pathways leading to adverse effects (4-7).

Drug development typically involves the screening of large libraries of lead compounds to identify those capable of binding to and signalling via a receptor. The vast array of lead compounds available requires high throughput screening (HTS) methods to enable identification of potential therapeutic compounds. Many of the HTS methods available for opioid receptors involve the expression of highly engineered proteins, require harvesting or lysing cells, and rely on a single endpoint measurement (8, 9). Here, we describe two simple, real-time assays reflecting a naturalistic coupling of MOPr to Gα and Gβγ proteins, using a proprietary membrane potential sensitive dye.

In mouse pituitary AtT-20 cells, heterologously expressed MOPr activates endogenous G- protein-gated inwardly rectifying potassium channels (GIRKs), via direct coupling of Gβγ subunits to the channels (10). The resulting membrane hyperpolarisation can be measured using a membrane potential sensitive dye. The extent of hyperpolarisation corresponds to the number of activated GIRK channels, and therefore closely reflects of the degree of

MOPr activation (11).

331 Inhibition of AC activity leading to a decrease in cAMP production is one of the hallmarks of MOPr activation and measurement of cAMP accumulation is frequently used to examine opioid potency and efficacy. Most cAMP accumulation assays rely on end-point measurements after significant incubation times to allow cAMP accumulation (9, 12,13), while real-time assays of cAMP accumulation such as the GloSensor assay (Promega), require transfection of sensor constructs with cAMP binding domains and use of specialized reagents. (14). The assay described here is a real-time, robust assay of AC inhibition in live CHO cells requiring minimal preparation and reagents. Stimulation of cAMP production Chinese hamster ovary (CHO) cells via the AC activator forskolin results in membrane hyperpolarisation, which is inhibited with the simultaneous addition of opioids (15).

Both assays described here have z-factors of 0.7, indicating sufficient robustness for HTS

(11, 15, 16). These assays offer an alternative approach for measuring MOPr activation by targeting naturalistic signalling pathways. The membrane potential assay is a rapid, reliable, and inexpensive method for identifying ligands that modulate Gα and Gβγ- mediated signalling and may be scaled up to enable HTS for novel opioid drugs.

2. Materials

1. FlexStation3 Plate Reader, running at 37oC. See Note 1.

2. Incubator with room air, at 37oC.

3. FLIPR Membrane Potential Dye (Molecular Devices) – blue rather than red. See Note

2.

4. Assay Buffer – consisting of (in mM): NaCl 145, HEPES 22, Na2HPO4 0.338,

NaHCO3 4.17, KH2PO4 0.441, MgSO4 0.407, MgCl2 0.493, CaCl2 1.26, Glucose 5.56,

pH 7.4, osmolarity 310-320

332 5. AtT-20 cells expressing MOPr (GIRK channel activation assay), or CHO-K1 cells

expressing MOPr (AC inhibition assay) at 90 % or more confluency in 75 cm2 tissue

culture flasks (1 flask per microplate)

6. Leibovitz’s L-15 media supplemented with 1% FBS, 100U penicillin/streptomycin ml-

1 and 15 mM glucose. See Note 3.

7. 96-well black-walled, sterile clear-bottomed microplates

8. 96-well clear, v-bottomed microplates

9. 96-well Black FlexStation pipette tips (Molecular Devices)

10. 8-channel multi-pipette

11. Troughs for holding cells and dye while using multichannel pipette.

12. Opioid ligands

13. Forskolin (AC inhibition assay)

3. Methods

1. On the day before the assay, harvest cells from single 75 cm2 flask, re-suspend the

cells obtained from each flask in 10 mL L-15 media, and plate in a volume of 90

µL/well in a black-walled 96-well plate using 8 channel pipettor. Incubate overnight at

37° C in ambient CO2. See Note 4.

2. Prepare the dye according to the manufacturer’s instructions in assay buffer. Prepared

dye can be stored at -80° C for several months. See Note 5.

333 3.1 GIRK Channel Activation Assay (AtT-20 cells)

1. Load the cells with 90 µL/well prepared membrane potential dye. Incubate for a

minimum of 45 minutes at 37° C in ambient CO2. Ensure the plate is uncovered for the

final 15 mins of incubation. See Notes 6 - 8.

2. Prepare drug solutions: Make up drugs in assay buffer at 10 x final concentration

desired. See Note 9. Load 200 µL/well drug solutions or vehicle into wells in v-bottom

96-well plate. See Notes 10 - 11.

Insert drug plate into appropriate FlexStation drawer and incubate for approx. 10 mins

to warm solutions to 37° C. See Note 12.

3. Set up the assay parameters as follows:

Read Mode: Fluorescence, bottom read

Excitation wavelength: 530

Emission wavelength: 565

Cutoff: Auto

Readings per well: 6

PMT: Medium

Run time: 300 secs

Interval: 2 secs

Select appropriate assay plate type and wells to read. See Note 13.

Compound transfer: Initial volume: 180 µL

Transfers: 1

Pipette Height: 190 µL

Volume: 20 µL

Rate: 2

334 Time point: 120 secs – this will allow a stable baseline to be established before drug addition.

Select appropriate compound plate.

Triturate: Select Assay plate, volume 20 µL, 3 cycles, pipette height 150 µL. Pipette

Tip Layout: Select columns of tips to be used corresponding to columns of wells to be

read. Tips may be re-used up to 3 times for replicates.

Select “Read”. See Note 14 - 17.

3.2 Adenylyl Cyclase Inhibition Assay (CHO cells)

1. Load the cells with 90 µL/well prepared membrane potential dye. Incubate for a

minimum of 60 minutes at 37° C in ambient CO2. See Note 6 – 8.

2. Prepare drug solutions:

Forskolin is added simultaneously with drug or vehicle. Make up drugs with forskolin

in assay buffer at 10 x final concentration desired, e.g. 3 µM FSK + 10 nM – 10 µM

DAMGO for a 1 nM – 1 µM DAMGO concentration response curve. See Note 9.

Load 200 µL/well drug solutions or vehicle into wells in v-bottom 96-well plate. See

Notes 10 - 11.

The FlexStation 3 reads column by column, so concentration response curves are best

set up within columns rather than across rows. Insert drug plate into appropriate

FlexStation drawer and incubate for approx. 10 mins to warm solutions to 37° C. See

Note 12.

3. Set up the assay parameters as follows:

Read Mode: Fluorescence, bottom read

Excitation wavelength: 530

335 Emission wavelength: 565

Cutoff: Auto

Readings per well: 6

PMT: Medium

Run time: 720 secs

Interval: 2 secs

Select appropriate assay plate type and wells to read. See Note 13.

Compound transfer:

Initial volume: 180 µL

Transfers: 1

Pipette Height: 190 µL

Volume: 20 µL

Rate: 2

Time point: 120 secs – this will allow a stable baseline to be established before drug addition.

Select appropriate compound plate

Triturate: Select Assay plate, volume 20 µL, 3 cycles, pipette height 150 µL Pipette

Tip Layout: Select columns of tips to be used corresponding to columns of wells to be read. Tips may be re-used up to 3 times for replicates.

Select “Read”. See Notes 13 – 15, 17 - 18.

336 700 RFU 300 nM DAMGO 600

500

400

300

200 Peak of membrane hyperpolarisation

100

0 060 120 180 240 300

Time (secs) Figure 1

Figure 1: Example trace of fluorescent signal in AtT-20-MOPr cells with the addition of 300 nM DAMGO at 120 sec. The rapid drop in fluorescent signal peaks 20 seconds after DAMGO addition, at which point measurements were taken.

700 300 nM Forskolin RFU 300 nM Forskolin + 1 uM DAMGO 600

500

400

300

200

100 Peak FSK response 0 0120 240 360 480 600 720 Time (secs) Figure 2

Figure 2: Example trace of the fluorescent signal in CHO-MOPr cells with the addition of 300 nM FSK alone or 300 nM FSK + 1 μM DAMGO. The maximum FSK-stimulated decrease in fluorescent signal occurred 5 min after addition, after which time the signal stabilized. Measurements for FSK alone and FSK + opioid were taken at this time-point.

337 4. Notes

1. We used a FlexStation 3 for this assay, however a similar plate reader with a fluidic

module could be used for the GIRK assay. The AC assay could potentially be

performed as an end-point assay on a static plate reader approximately 5 mins after

drug treatment, providing a 300 nM FSK control is included in each column.

2. The membrane potential-sensitive dye is available in a blue or red formulation. We

found blue dye to give the largest signal window for our cell lines, however, this may

vary. To ensure optimal results both blue and red dyes should be trialled, particularly

if adapting the assays for other cell types.

3. Serum-starving the cells synchronises the cell cycle phase and helps to increase

reproducibility of the results. Serum starving also prevents cell overgrowing and

forming multiple layers in wells.

4. For the cell lines used in these assays, the use of polylysine coated plates was not

necessary. We have found that experiments with other AtT-20 lines, such as those

expressing the cannabinoid-CB1 receptor, require polylysine coated plates to prevent

cells detaching.

5. Membrane potential dye (blue) can be successfully used at half the concentration

suggested by the manufacturer for these assays with no appreciable difference in

signal, however we observed decreased stability of half-strength dye with longer

assays (> 20 min). We have not tested red membrane potential dye at half-

concentration.

6. Although the manufacturer’s protocol recommends an incubation time of 30 mins with

dye, we found that when readings were taken after 30 mins incubation the baseline

steadily increased over time, indicating incomplete uptake of dye into cells. An

incubation of 45 mins was sufficient to achieve a flat baseline for AtT-20 cells, a

338 minimum of 60 mins was necessary for CHO cells. This may vary between cell lines

and assay conditions. Incubate for sufficient time period to achieve a flat baseline.

7. For the GIRK channel activation assay, an entire plate may be read in approximately

one hour, so it is practical to load an entire plate with dye at once with no deterioration

of signal. For longer assays such as the AC inhibition assay, dye loading can be

staggered so that cells are not loaded with dye for an extended period of time. In this

case, reading of half a plate takes approximately 80 mins, so load half a plate with dye,

incubate for 80 mins, then immediately prior to reading load the second half of the

plate with dye and cover with parafilm, allowing dye to load in the second half of the

plate while the first half is being read.

8. Where possible, the assay plate should be incubated in the FlexStation 3 while the dye

is loading to minimize temperature changes when transferring between the incubator

and Flexstation (in our lab these are located in different rooms). When this is not

possible, the assay plate may be incubated in an incubator, and transferred to the

FlexStation for the last 15 mins of incubation, uncovered.

9. When making up drug solutions, keep the concentration of any solvents used (e.g.

DMSO or ethanol) constant. Also include the same concentration of solvent in the

vehicle blank.

10. There should be a minimum of 80 µL of compound in the wells of the compound plate

to ensure consistency in compound transfer volume. Compound may be taken from the

same column for replicates until volume in well reaches 60 μL.

11. For relatively short assays, no changes in response over time were detected when

reading an entire plate, i.e. there appears to be little effect of evaporation of drug or

dye from compound and assay plates. For the AC inhibition assay, the compound plate

should be only be loaded with the drugs required for the portion of the plate being

339 read. Load fresh compounds when beginning a new read for the second half of the

plate.

12. Some compounds, for example somatostatin, may be unstable for extended time

periods at 37°C. In this case, drugs should be kept on ice and loaded into compound

plate just prior to reading.

13. For short assays, providing you are not using any unstable drugs, you can select and

read the entire plate. Otherwise, select only the columns you will read before loading

fresh drug solutions. Note that when using SoftMax Pro with the FlexStation 3, only

consecutive columns can be read in a single run.

14. For this assay, on our Flexstation, a baseline of between 600 – 1200 RFU is optimal.

These values may vary between machines, but a low baseline does not give a

sufficient window for the decreased fluorescent signal associated with cellular

hyperpolarisation, and a high baseline can lead to unpredictable results for a number

of reasons including too many cells per well, cells which are excessively depolarized

or a double addition of dye. More than a single layer of cells can lead to problems

with drug access and also loss of cells when drug is added, which leads to a large and

immediate drop in signal.

15. The background signals for these assays are low – cells without dye usually read

approximately 20-30 RFU (< 5% of the optimal signal), while L-15 plus dye without

cells gives a negligible reading. We do not routinely correct for these background

signals, but they should checked occasionally by reading a few wells without dye, with

and without cells.

16. GIRK Activation Assay Data Analysis: For this assay, we calculated GIRK activation

as a percentage decrease in fluorescence from baseline. We calculated the mean of the

readings from 30 sec before drug addition, ensuring that the baseline is flat and stable.

The peak response occurs rapidly after drug addition, usually within 20 secs (see Fig

340 1). We took the mean of the lowest reading from the peak response and 2 readings

either side, calculating the percentage difference between this value and the mean

baseline. We usually include a buffer/solvent blank in each row and subtract the

change produced by the 20 µl addition of buffer from all the experimental data. This

change is usually less than 5%.

17. If you want to normalize data between assays, the hyperpolarisation produced by

activation of endogenous SST receptors in AtT-20 cells can be used, alternatively

nigericin (1 µM), a K-selective ionophore, can be used to determine the fluorescence

change asscoiated with hyperpolarisation of the cells to EK. Both somatostatin and

nigericin produce decreases in fluorescence greater than the maximum produced by

opioids.

18. AC Inhibition Data Analysis: For this assay, we calculated AC inhibition as a

percentage inhibition of the FSK response. The FSK response was calculated as the

maximum % decrease in fluorescence from baseline after addition of FSK alone. We

calculated the mean of the readings from 30 sec before drug addition, ensuring that the

baseline is flat and stable. The peak response of FSK occurs approximately 5 min after

drug addition (see Fig 2). We took the mean of the lowest reading from the peak

response and 2 readings either side, calculating the percentage difference between this

value and the mean baseline for the maximum FSK response. We took readings at the

same timepoint when FSK was added with opioid, and calculated the difference

between the FSK response and the FSK + opioid response, and expressed as a

percentage inhibition of FSK response. We also included a buffer/solvent blank in

each row and subtracted the change produced by the 20 µl addition of buffer

containing the solvent for FSK from the experimental data. This change is usually less

than 5%.

341 References

1. Matthes, H. W. D., Maldonado, R., Simonin, F., et al. (1996). Loss of Morphine-

Induced Analgesia, Reward Effect and Withdrawal Symptoms in Mice Lacking the μ-

Opioid-Receptor Gene. Nature 383, 819-823

2. Kenakin, T, Miller, L.J. (2010). Seven transmembrane receptors as shapeshifting

proteins: the impact of allosteric modulation and functional selectivity on new drug

discovery. Pharmacol Rev 62, 265-304

3. Pinyero, G., Archer-Lahlou, E. (2007). Ligand-specific receptor states: implications

for opiate receptor signalling and regulation. Cell Signal 19, 8-19

4. DeWire, S.M., Yamashita, D.S., Rominger D.H. et al. (2013). A G protein-biased

ligand at the μ-opioid receptor is potently analgesic with reduced gastrointestinal and

respiratory dysfunction compared with morphine. J Pharmacol Exp Ther 344, 708-717

5. Kenakin, T., Christopoulos, A. (2013). Signalling bias in new drug discovery:

detection, quantification and therapeutic impact. Nat Rev Drug Discov 12, 205-216

6. Pradhan, A.A., Smith, M.L., Kieffer, B.L. et al. (2012). Ligand-directed signalling

within the opioid receptor family. Br J Pharmacol 167, 960-969

7. Zheng, H., Loh, H.H., Law, P.Y. (2013) Post-translation modification of G protein-

copuled receptor in relationship to biased agonism. Methods Enzymol 522, 391-408

8. Coward, P., Chan, S.D.H., Wada H.G. et al. (1999). Chimeric G-proteins allow a high-

throughput signalling assay of Gi-coupled receptors. Anal. Biochem. 270, 242-248

9. Hill, S.J., Williams, C., May, L.T. (2010). Insights into GPCR pharmacology from the

measurement of changes in intracellular cyclic AMP: Advantages and pitfalls of

differing methodologies. Br J Pharmacol 161, 1266-1275

10. Celver, J., Xu, M., Jin, W. et al. (2004) Distinct Domains of the μ-opioid receptor

control uncoupling and internalisation. Mol Pharmacol 65, 528-537

11. Knapman, A., Santiago, M., Du, Y.P et al. (2013) A continuous, fluorescence-based

342 assay of μ-opioid receptor activation in AtT-20 cells. J Biomol Screen 18, 269-276

12. Gabriel, D., Vernier, M., Pfeifer, M.J. et al. (2003) High throughput screening

technologies for direct cyclic AMP measurement. Assay Drug Dev. Technol 1, 291-

303

13. Williams, C. (2004). cAMP detection methods in HTS: Selecting the best from the

rest. Nat Rev Drug Discov 3, 125-135

14. Binkowski, B.F., Fan, F., Wood, K.V. (2011). Luminescent biosensors for real time

monitoring of intracellular cAMP. Methods Mol Biol 756, 263-271

15. Knapman A., Abodadie F., McIntyre P. et al. (2014). A real-time, fluorescence-based

assay for measuring μ-opioid receptor modulation of adenylyl cyclase activity in

Chinese hamster ovary cells. J Biomol Screen 19, 223-231

16. Zhang, J., Chung, T.D.Y., Oldenburg K.R. (1999) A simple statistical parameter for

use in evaluation and validation of high throughput screening assays. J Biomol Screen

4, 67-73

343